Board Game Behavior: An Analysis of Demographics and Gameplay Preferences¶
Data Processing and Modeling¶
In [1]:
!python -m pip install pandas
!python -m pip install openpyxl
!python -m pip install seaborn
!python -m pip install scikit-learn
Collecting pandas
Downloading pandas-2.0.3-cp38-cp38-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (18 kB)
Requirement already satisfied: python-dateutil>=2.8.2 in /opt/hostedtoolcache/Python/3.8.18/x64/lib/python3.8/site-packages (from pandas) (2.8.2)
Requirement already satisfied: pytz>=2020.1 in /opt/hostedtoolcache/Python/3.8.18/x64/lib/python3.8/site-packages (from pandas) (2023.3.post1)
Collecting tzdata>=2022.1 (from pandas)
Downloading tzdata-2023.3-py2.py3-none-any.whl (341 kB)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 341.8/341.8 kB 4.7 MB/s eta 0:00:00
Requirement already satisfied: numpy>=1.20.3 in /opt/hostedtoolcache/Python/3.8.18/x64/lib/python3.8/site-packages (from pandas) (1.24.4)
Requirement already satisfied: six>=1.5 in /opt/hostedtoolcache/Python/3.8.18/x64/lib/python3.8/site-packages (from python-dateutil>=2.8.2->pandas) (1.16.0)
Downloading pandas-2.0.3-cp38-cp38-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (12.4 MB)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 12.4/12.4 MB 21.9 MB/s eta 0:00:00
Installing collected packages: tzdata, pandas
Successfully installed pandas-2.0.3 tzdata-2023.3
[notice] A new release of pip is available: 23.0.1 -> 23.3.1
[notice] To update, run: pip install --upgrade pip
Collecting openpyxl
Downloading openpyxl-3.1.2-py2.py3-none-any.whl (249 kB)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 250.0/250.0 kB 6.3 MB/s eta 0:00:00
Collecting et-xmlfile (from openpyxl)
Downloading et_xmlfile-1.1.0-py3-none-any.whl (4.7 kB)
Installing collected packages: et-xmlfile, openpyxl
Successfully installed et-xmlfile-1.1.0 openpyxl-3.1.2
[notice] A new release of pip is available: 23.0.1 -> 23.3.1
[notice] To update, run: pip install --upgrade pip
Collecting seaborn
Downloading seaborn-0.13.0-py3-none-any.whl.metadata (5.3 kB)
Requirement already satisfied: numpy!=1.24.0,>=1.20 in /opt/hostedtoolcache/Python/3.8.18/x64/lib/python3.8/site-packages (from seaborn) (1.24.4)
Requirement already satisfied: pandas>=1.2 in /opt/hostedtoolcache/Python/3.8.18/x64/lib/python3.8/site-packages (from seaborn) (2.0.3)
Requirement already satisfied: matplotlib!=3.6.1,>=3.3 in /opt/hostedtoolcache/Python/3.8.18/x64/lib/python3.8/site-packages (from seaborn) (3.7.4)
Requirement already satisfied: contourpy>=1.0.1 in /opt/hostedtoolcache/Python/3.8.18/x64/lib/python3.8/site-packages (from matplotlib!=3.6.1,>=3.3->seaborn) (1.1.1)
Requirement already satisfied: cycler>=0.10 in /opt/hostedtoolcache/Python/3.8.18/x64/lib/python3.8/site-packages (from matplotlib!=3.6.1,>=3.3->seaborn) (0.12.1)
Requirement already satisfied: fonttools>=4.22.0 in /opt/hostedtoolcache/Python/3.8.18/x64/lib/python3.8/site-packages (from matplotlib!=3.6.1,>=3.3->seaborn) (4.46.0)
Requirement already satisfied: kiwisolver>=1.0.1 in /opt/hostedtoolcache/Python/3.8.18/x64/lib/python3.8/site-packages (from matplotlib!=3.6.1,>=3.3->seaborn) (1.4.5)
Requirement already satisfied: packaging>=20.0 in /opt/hostedtoolcache/Python/3.8.18/x64/lib/python3.8/site-packages (from matplotlib!=3.6.1,>=3.3->seaborn) (23.2)
Requirement already satisfied: pillow>=6.2.0 in /opt/hostedtoolcache/Python/3.8.18/x64/lib/python3.8/site-packages (from matplotlib!=3.6.1,>=3.3->seaborn) (10.1.0)
Requirement already satisfied: pyparsing>=2.3.1 in /opt/hostedtoolcache/Python/3.8.18/x64/lib/python3.8/site-packages (from matplotlib!=3.6.1,>=3.3->seaborn) (3.1.1)
Requirement already satisfied: python-dateutil>=2.7 in /opt/hostedtoolcache/Python/3.8.18/x64/lib/python3.8/site-packages (from matplotlib!=3.6.1,>=3.3->seaborn) (2.8.2)
Requirement already satisfied: importlib-resources>=3.2.0 in /opt/hostedtoolcache/Python/3.8.18/x64/lib/python3.8/site-packages (from matplotlib!=3.6.1,>=3.3->seaborn) (6.1.1)
Requirement already satisfied: pytz>=2020.1 in /opt/hostedtoolcache/Python/3.8.18/x64/lib/python3.8/site-packages (from pandas>=1.2->seaborn) (2023.3.post1)
Requirement already satisfied: tzdata>=2022.1 in /opt/hostedtoolcache/Python/3.8.18/x64/lib/python3.8/site-packages (from pandas>=1.2->seaborn) (2023.3)
Requirement already satisfied: zipp>=3.1.0 in /opt/hostedtoolcache/Python/3.8.18/x64/lib/python3.8/site-packages (from importlib-resources>=3.2.0->matplotlib!=3.6.1,>=3.3->seaborn) (3.17.0)
Requirement already satisfied: six>=1.5 in /opt/hostedtoolcache/Python/3.8.18/x64/lib/python3.8/site-packages (from python-dateutil>=2.7->matplotlib!=3.6.1,>=3.3->seaborn) (1.16.0)
Downloading seaborn-0.13.0-py3-none-any.whl (294 kB)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 294.6/294.6 kB 5.1 MB/s eta 0:00:00
Installing collected packages: seaborn
Successfully installed seaborn-0.13.0
[notice] A new release of pip is available: 23.0.1 -> 23.3.1
[notice] To update, run: pip install --upgrade pip
Collecting scikit-learn
Downloading scikit_learn-1.3.2-cp38-cp38-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (11 kB)
Requirement already satisfied: numpy<2.0,>=1.17.3 in /opt/hostedtoolcache/Python/3.8.18/x64/lib/python3.8/site-packages (from scikit-learn) (1.24.4)
Collecting scipy>=1.5.0 (from scikit-learn)
Downloading scipy-1.10.1-cp38-cp38-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (34.5 MB)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 34.5/34.5 MB 69.8 MB/s eta 0:00:00
Collecting joblib>=1.1.1 (from scikit-learn)
Downloading joblib-1.3.2-py3-none-any.whl.metadata (5.4 kB)
Collecting threadpoolctl>=2.0.0 (from scikit-learn)
Downloading threadpoolctl-3.2.0-py3-none-any.whl.metadata (10.0 kB)
Downloading scikit_learn-1.3.2-cp38-cp38-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (11.1 MB)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 11.1/11.1 MB 106.8 MB/s eta 0:00:00
Downloading joblib-1.3.2-py3-none-any.whl (302 kB)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 302.2/302.2 kB 60.9 MB/s eta 0:00:00
Downloading threadpoolctl-3.2.0-py3-none-any.whl (15 kB)
Installing collected packages: threadpoolctl, scipy, joblib, scikit-learn
Successfully installed joblib-1.3.2 scikit-learn-1.3.2 scipy-1.10.1 threadpoolctl-3.2.0
[notice] A new release of pip is available: 23.0.1 -> 23.3.1
[notice] To update, run: pip install --upgrade pip
Data Cleaning¶
In [2]:
import pandas as pd
import numpy as np
raw_df = pd.read_excel("datasets/rawdata.xlsx",keep_default_na=False)
raw_df.info()
<class 'pandas.core.frame.DataFrame'>
RangeIndex: 244 entries, 0 to 243
Data columns (total 22 columns):
# Column Non-Null Count Dtype
--- ------ -------------- -----
0 ID 244 non-null int64
1 Start time 244 non-null datetime64[ns]
2 Completion time 244 non-null datetime64[ns]
3 Email 244 non-null object
4 Name 244 non-null object
5 Last modified time 244 non-null object
6 I am a: 244 non-null object
7 What is your Gender Identity? 244 non-null object
8 What is your Age?
(this field may remain blank) 244 non-null object
9 What is your Race? 244 non-null object
10 What is your current Employment Status? 244 non-null object
11 Do you wear glasses or contact lenses for vision correction? 244 non-null object
12 What best describes your current religious or spiritual beliefs, if any?
244 non-null object
13 What is your Area of Study or Major?
(this field may remain blank)
244 non-null object
14 Do you enjoy playing board games, card games, or similar tabletop games? 244 non-null object
15 How many board/card games do you own?
(All non-roleplay table top games are included, i.e., chess, playing cards, Uno, Magic: The Gathering, etc.) 244 non-null object
16 About how frequently do you play board/card games? 244 non-null object
17 Select what best describes your engagement and style when it comes to board and card games? 244 non-null object
18 I prefer games that include elements of:
(select all that you prefer) 244 non-null object
19 List up to 15 board/card games that you enjoy playing.
(abide by the format: "Game 1", "Game 2", "Game 3")
(this field may remain blank)
244 non-null object
20 On a scale from 1 to 10, how much do you enjoy playing chess, with 1 being not enjoyable at all and 10 being extremely enjoyable? 244 non-null object
21 Select your all genres of board/card games you enjoy.
(if your preferred genre(s) is/are not listed, please list them in the "other" selection) 244 non-null object
dtypes: datetime64[ns](2), int64(1), object(19)
memory usage: 42.1+ KB
Drop Metadata¶
In [3]:
dropped_df = raw_df.drop(columns=["ID","Start time","Completion time","Email","Name","Last modified time"])
dropped_df[:0]
Out[3]:
| I am a: | What is your Gender Identity? | What is your Age?\n(this field may remain blank) | What is your Race? | What is your current Employment Status? | Do you wear glasses or contact lenses for vision correction? | What best describes your current religious or spiritual beliefs, if any?\n | What is your Area of Study or Major? \n(this field may remain blank)\n | Do you enjoy playing board games, card games, or similar tabletop games? | How many board/card games do you own? \n(All non-roleplay table top games are included, i.e., chess, playing cards, Uno, Magic: The Gathering, etc.) | About how frequently do you play board/card games? | Select what best describes your engagement and style when it comes to board and card games? | I prefer games that include elements of:\n(select all that you prefer) | List up to 15 board/card games that you enjoy playing. \n(abide by the format: "Game 1", "Game 2", "Game 3")\n(this field may remain blank)\n | On a scale from 1 to 10, how much do you enjoy playing chess, with 1 being not enjoyable at all and 10 being extremely enjoyable? | Select your all genres of board/card games you enjoy.\n(if your preferred genre(s) is/are not listed, please list them in the "other" selection) |
|---|
Rename Raw Fields from Question to Concise Description¶
In [4]:
renamed_df = dropped_df.rename(columns={
"I am a:": "WWUStatus",
"What is your Gender Identity?": "Gender",
"What is your Age?\n(this field may remain blank)": "Age",
"What is your Race?": "Race(s)",
"What is your current Employment Status?": "EmploymentStatus",
"Do you wear glasses or contact lenses for vision correction?": "Vision",
"What best describes your current religious or spiritual beliefs, if any?\n": "Religiosity",
"What is your Area of Study or Major? \n(this field may remain blank)\n": "AOS",
"Do you enjoy playing board games, card games, or similar tabletop games?": "EnjoysBoardGames",
"How many board/card games do you own? \n(All non-roleplay table top games are included, i.e., chess, playing cards, Uno, Magic: The Gathering, etc.)": "BoardGamesOwned",
"About how frequently do you play board/card games?": "FrequencyOfPlay",
"Select what best describes your engagement and style when it comes to board and card games? ": "Style",
"I prefer games that include elements of:\n(select all that you prefer)": "PreferredElements",
"List up to 15 board/card games that you enjoy playing. \n(abide by the format: \"Game 1\", \"Game 2\", \"Game 3\")\n(this field may remain blank)\n": "EnjoyedBoardGames",
"On a scale from 1 to 10, how much do you enjoy playing chess, with 1 being not enjoyable at all and 10 being extremely enjoyable?": "ChessRating",
"Select your all genres of board/card games you enjoy.\n(if your preferred genre(s) is/are not listed, please list them in the \"other\" selection)": "EnjoyedGenres"
})
renamed_df[:0]
Out[4]:
| WWUStatus | Gender | Age | Race(s) | EmploymentStatus | Vision | Religiosity | AOS | EnjoysBoardGames | BoardGamesOwned | FrequencyOfPlay | Style | PreferredElements | EnjoyedBoardGames | ChessRating | EnjoyedGenres |
|---|
Multiselect Binary/Boolean Features¶
Race(s)¶
In [5]:
renamed_df["Race(s)"].value_counts()
Out[5]:
Race(s) White; 185 Asian; 14 Hispanic or Latino; 9 White;Asian; 6 White;Hispanic or Latino; 5 Black or African American;White; 4 White;Black or African American; 3 White;Native Hawaiian or Pacific Islander; 2 Black or African American; 2 Asian;Hispanic or Latino; 2 Black or African American;White;Asian; 1 Black or African American;Hispanic or Latino; 1 White;Native American or American Indian; 1 Multiracial; 1 Mixed ethnicity; 1 Native American or American Indian; 1 Asian;Filipino; 1 Hispanic or Latino;White; 1 Asian;White; 1 White;Black or African American;Asian; 1 Prefer not to say; 1 Native American or American Indian;White; 1 Name: count, dtype: int64
In [6]:
raceClean_df = renamed_df.copy()
unique_races = set()
for elements in raceClean_df["Race(s)"].dropna():
unique_races.update(elements.split(";"))
unique_races = list(unique_races)
unique_races.sort()
In [7]:
# Create a new column for each unique element with binary values (0 or 1 or None)
for element in unique_races:
raceClean_df[(element + "IsRace").replace(" ", "")] = raceClean_df["Race(s)"].apply(
lambda x: int(element in x) if pd.notna(x) else None
)
raceClean_df = raceClean_df.drop(columns={"Race(s)","IsRace"})
PreferredElements¶
In [8]:
raceClean_df["PreferredElements"].value_counts()
Out[8]:
PreferredElements
Conflict/Competition;Strategy; 7
Cooperation;Conflict/Competition;Luck;Strategy;Social Deduction/Hidden Role;Heavy/Immersive Theming;Puzzle-Solving;Trivia;Party/Low-Stakes; 6
Party/Low-Stakes; 4
Conflict/Competition;Strategy;Trivia; 4
Cooperation;Conflict/Competition;Luck;Strategy;Puzzle-Solving;Party/Low-Stakes; 3
..
Cooperation;Conflict/Competition;Strategy;Luck; 1
Cooperation;Strategy;Luck;Conflict/Competition;Social Deduction/Hidden Role;Trivia;Party/Low-Stakes; 1
Luck;Strategy;Conflict/Competition;Party/Low-Stakes;Trivia; 1
Cooperation;Luck;Conflict/Competition; 1
Social Deduction/Hidden Role;Party/Low-Stakes; 1
Name: count, Length: 202, dtype: int64
In [9]:
preferredElementsClean_df = raceClean_df.copy()
unique_preferred_elements = set()
for elements in preferredElementsClean_df["PreferredElements"].dropna():
unique_preferred_elements.update(elements.split(";"))
unique_preferred_elements
unique_preferred_elements = list(unique_preferred_elements)
unique_preferred_elements.sort()
In [10]:
# Create a new column for each unique element with binary values (0 or 1 or None)
for element in unique_preferred_elements:
preferredElementsClean_df[(element + "IsPreferredElement").replace(" ", "")] = preferredElementsClean_df["PreferredElements"].apply(
lambda x: int(element in x) if pd.notna(x) else None
)
preferredElementsClean_df = preferredElementsClean_df.drop(columns={"PreferredElements","IsPreferredElement"})
EnjoyedGenres¶
In [11]:
enjoyedGenresClean_df = preferredElementsClean_df.copy()
unique_enjoyed_genres = set()
for elements in enjoyedGenresClean_df["EnjoyedGenres"].dropna():
unique_enjoyed_genres.update(elements.split(";"))
unique_enjoyed_genres
unique_enjoyed_genres = list(unique_enjoyed_genres)
unique_enjoyed_genres.sort()
In [12]:
# Create a new column for each unique element with binary values (0 or 1 or None)
for element in unique_enjoyed_genres:
enjoyedGenresClean_df[(element + "IsEnjoyedGenre").replace(" ", "")] = enjoyedGenresClean_df["EnjoyedGenres"].apply(
lambda x: int(element in x) if pd.notna(x) else None
)
enjoyedGenresClean_df = enjoyedGenresClean_df.drop(columns={"EnjoyedGenres","IsEnjoyedGenre"})
In [13]:
enjoyedGenresClean_df.columns.tolist()
Out[13]:
['WWUStatus', 'Gender', 'Age', 'EmploymentStatus', 'Vision', 'Religiosity', 'AOS', 'EnjoysBoardGames', 'BoardGamesOwned', 'FrequencyOfPlay', 'Style', 'EnjoyedBoardGames', 'ChessRating', 'AsianIsRace', 'BlackorAfricanAmericanIsRace', 'FilipinoIsRace', 'HispanicorLatinoIsRace', 'MixedethnicityIsRace', 'MultiracialIsRace', 'NativeAmericanorAmericanIndianIsRace', 'NativeHawaiianorPacificIslanderIsRace', 'PrefernottosayIsRace', 'WhiteIsRace', 'Conflict/CompetitionIsPreferredElement', 'CooperationIsPreferredElement', 'Heavy/ImmersiveThemingIsPreferredElement', 'LuckIsPreferredElement', 'Party/Low-StakesIsPreferredElement', 'Puzzle-SolvingIsPreferredElement', 'SocialDeduction/HiddenRoleIsPreferredElement', 'StrategyIsPreferredElement', 'TriviaIsPreferredElement', 'AbstractStrategyIsEnjoyedGenre', 'AdventureIsEnjoyedGenre', 'AnimalsIsEnjoyedGenre', 'AuctionIsEnjoyedGenre', 'CardIsEnjoyedGenre', 'CardDraftingIsEnjoyedGenre', 'CivilizationIsEnjoyedGenre', 'Cooperative\xa0IsEnjoyedGenre', 'Deck-Building\xa0IsEnjoyedGenre', 'DeductionIsEnjoyedGenre', 'EconomicIsEnjoyedGenre', 'EducationalIsEnjoyedGenre', 'ExplorationIsEnjoyedGenre', 'FantasyIsEnjoyedGenre', 'FarmingIsEnjoyedGenre', 'FightingIsEnjoyedGenre', 'HorrorIsEnjoyedGenre', 'LuckIsEnjoyedGenre', 'MedievalIsEnjoyedGenre', 'MemoryIsEnjoyedGenre', 'MiniaturesIsEnjoyedGenre', 'Party\xa0IsEnjoyedGenre', 'PiratesIsEnjoyedGenre', 'PoliticalIsEnjoyedGenre', 'PuzzleIsEnjoyedGenre', 'RacingIsEnjoyedGenre', 'Role-Playing\xa0IsEnjoyedGenre', 'RollandMove\xa0IsEnjoyedGenre', 'ScienceFictionIsEnjoyedGenre', 'SocialDeduction/HiddenRoleIsEnjoyedGenre', 'SportsIsEnjoyedGenre', 'StrategyIsEnjoyedGenre', 'TerritoryBuildingIsEnjoyedGenre', 'Tile-Laying\xa0IsEnjoyedGenre', 'TrainsIsEnjoyedGenre', 'TransportationIsEnjoyedGenre', 'TravelIsEnjoyedGenre', 'TriviaIsEnjoyedGenre', 'War\xa0IsEnjoyedGenre', 'Word\xa0IsEnjoyedGenre', 'WorkerPlacementIsEnjoyedGenre', 'WorldWarIIIsEnjoyedGenre', 'ZombiesIsEnjoyedGenre']
In [14]:
reformatColumns_df = enjoyedGenresClean_df.rename(columns=lambda x: x.replace("\xa0", ""))
reformatColumns_df.columns.tolist()
Out[14]:
['WWUStatus', 'Gender', 'Age', 'EmploymentStatus', 'Vision', 'Religiosity', 'AOS', 'EnjoysBoardGames', 'BoardGamesOwned', 'FrequencyOfPlay', 'Style', 'EnjoyedBoardGames', 'ChessRating', 'AsianIsRace', 'BlackorAfricanAmericanIsRace', 'FilipinoIsRace', 'HispanicorLatinoIsRace', 'MixedethnicityIsRace', 'MultiracialIsRace', 'NativeAmericanorAmericanIndianIsRace', 'NativeHawaiianorPacificIslanderIsRace', 'PrefernottosayIsRace', 'WhiteIsRace', 'Conflict/CompetitionIsPreferredElement', 'CooperationIsPreferredElement', 'Heavy/ImmersiveThemingIsPreferredElement', 'LuckIsPreferredElement', 'Party/Low-StakesIsPreferredElement', 'Puzzle-SolvingIsPreferredElement', 'SocialDeduction/HiddenRoleIsPreferredElement', 'StrategyIsPreferredElement', 'TriviaIsPreferredElement', 'AbstractStrategyIsEnjoyedGenre', 'AdventureIsEnjoyedGenre', 'AnimalsIsEnjoyedGenre', 'AuctionIsEnjoyedGenre', 'CardIsEnjoyedGenre', 'CardDraftingIsEnjoyedGenre', 'CivilizationIsEnjoyedGenre', 'CooperativeIsEnjoyedGenre', 'Deck-BuildingIsEnjoyedGenre', 'DeductionIsEnjoyedGenre', 'EconomicIsEnjoyedGenre', 'EducationalIsEnjoyedGenre', 'ExplorationIsEnjoyedGenre', 'FantasyIsEnjoyedGenre', 'FarmingIsEnjoyedGenre', 'FightingIsEnjoyedGenre', 'HorrorIsEnjoyedGenre', 'LuckIsEnjoyedGenre', 'MedievalIsEnjoyedGenre', 'MemoryIsEnjoyedGenre', 'MiniaturesIsEnjoyedGenre', 'PartyIsEnjoyedGenre', 'PiratesIsEnjoyedGenre', 'PoliticalIsEnjoyedGenre', 'PuzzleIsEnjoyedGenre', 'RacingIsEnjoyedGenre', 'Role-PlayingIsEnjoyedGenre', 'RollandMoveIsEnjoyedGenre', 'ScienceFictionIsEnjoyedGenre', 'SocialDeduction/HiddenRoleIsEnjoyedGenre', 'SportsIsEnjoyedGenre', 'StrategyIsEnjoyedGenre', 'TerritoryBuildingIsEnjoyedGenre', 'Tile-LayingIsEnjoyedGenre', 'TrainsIsEnjoyedGenre', 'TransportationIsEnjoyedGenre', 'TravelIsEnjoyedGenre', 'TriviaIsEnjoyedGenre', 'WarIsEnjoyedGenre', 'WordIsEnjoyedGenre', 'WorkerPlacementIsEnjoyedGenre', 'WorldWarIIIsEnjoyedGenre', 'ZombiesIsEnjoyedGenre']
AOS (manual string bucketing)¶
In [15]:
aos_df = reformatColumns_df
aos_df["AOS"].str.lower().unique()
Out[15]:
array(['computer science ', 'data science ', 'data science', 'eece',
'electrical and computer engineering ', 'statistics',
'environmental studies', 'applied mathematics', 'chemistry', 'rml',
'political science', 'elementary education', 'english',
'music education', 'n/a', 'art', 'psychology', '',
'psych (probably)', 'music', 'environmental science - toxicology',
'history/museum studies', 'elementary ed',
'environmental science ', 'mathematics ', 'business', 'biochem ',
'vocal performance ', 'secondary education ', 'business ',
'linguistics', 'history', 'bio/anth', 'mathematics',
'marine biology ', 'environmental science',
'communication disorders', 'engineering ', 'biochem',
'kinesiology', 'economics and mathematics',
'music education and german', 'art p-12',
'chemistry either organic or inorganic', 'math',
'electrical engineering',
'undecided but leaning towards engineering', 'medicine ',
'rec management ', 'economics ', 'geology', 'visual journalism ',
'environmental studies ', 'biology/math',
'behavioral neuroscience', 'electrical engineering ',
'computer science', 'geology (paleoclimate)',
'marine biology and theater production', 'anthropology',
'biology ', 'management information systems ', 'marine bio',
'history/holocaust & genocide studies', 'sped & eled',
'visual journalism', 'anthropology, communication studies',
'theatre', 'studio art',
'urban planning and sustainable development',
'urban planning and sustainable development ', 'history ',
'art and design ', 'kinesiology ', 'spanish ', 'biochemistry ',
'art studio', 'art ed', 'comm', 'early childhood education ',
'creative writing', 'neuroscience ', 'marine science ',
'marketing ', 'behavioral neuroscience ', 'pre nursing ',
'engineering', 'graphic design', 'undecided',
'english literature with a teaching emphasis',
'political science ', 'international business ',
'communication studies', 'dance',
'narrative and folklore studies (fairhaven major) ', 'psychology ',
'anthropology ', 'pre med and psychology ', 'biology',
'education and public relations', 'economics/mathematics',
'communications', 'art studio (ba), art history',
'elementary education ', 'archaeology ', 'theatre/education',
'marketing', 'business and sustainability ', 'biochemistry',
'environmental studies: eco-social justice and education emphasis',
'education ', 'education', 'envs ',
'mathematics secondary education', 'music composition',
'sociology ', 'stem', 'linguistics ', 'fairhaven', 'fairhaven ',
'behavioural neuroscience', 'english lit',
'food equity and sustainable agriculture ',
'art history and museum studies', 'japanese language ',
'graphic design and marketing ', 'music performance major',
'environment studies', 'business or elementary education ',
'marine and coastal science',
'undeclared, strongly thinking about history ', 'public health',
'energy policy and management ', 'undeclared', 'fine arts',
'undecided ', 'english, history of culture ',
'psychology and elementary education ',
'communication science and disordwrs', 'anthropology/history',
'special education and elementary education ', 'ibus',
'energy science', 'politics/philosophy/economics', 'studio art ',
'history/social studies', 'energy'], dtype=object)
In [16]:
manual_mapping_aos = {
"computer science": "STEM",
"data science": "STEM",
"eece": "STEM",
"electrical and computer engineering": "STEM",
"statistics": "STEM",
"environmental studies": "STEM",
"applied mathematics": "STEM",
"chemistry": "STEM",
"rml": "Other",
"political science": "Social Studies",
"elementary education": "Education",
"english": "Arts & Humanities",
"music education": "Arts & Humanities",
"nan": "Other",
"art": "Arts & Humanities",
"psychology": "Social Studies",
"psych (probably)": "Social Studies",
"music": "Arts & Humanities",
"environmental science - toxicology": "STEM",
"history/museum studies": "Arts & Humanities",
"elementary ed": "Education",
"environmental science": "STEM",
"mathematics": "STEM",
"business": "Business",
"biochem": "STEM",
"vocal performance": "Arts & Humanities",
"secondary education": "Education",
"linguistics": "Arts & Humanities",
"history": "Arts & Humanities",
"bio/anth": "STEM",
"marine biology": "STEM",
"communication disorders": "Health & Medicine",
"engineering": "STEM",
"kinesiology": "Health & Medicine",
"economics and mathematics": "STEM",
"music education and german": "Arts & Humanities",
"art p-12": "Arts & Humanities",
"chemistry either organic or inorganic": "STEM",
"math": "STEM",
"electrical engineering": "STEM",
"undecided but leaning towards engineering": "Unknown", # Assuming lean towards STEM, but no exact match
"medicine": "Health & Medicine",
"rec management": "Other", # Assuming Recreation Management
"economics": "Social Studies", # Close to "economics and mathematics", but economics is often considered Social Studies
"geology": "STEM",
"geology (paleoclimate)": "STEM",
"visual journalism": "Arts & Humanities",
"biology/math": "STEM", # Combination of two STEM fields
"behavioral neuroscience": "STEM", # Close to "psychology" which is Social Studies, but has a heavy STEM component
"marine biology and theater production": "STEM", # Marine biology is STEM, theater production could be Arts, but STEM is the primary
"anthropology": "Social Studies",
"biology": "STEM",
"management information systems": "Business",
"marine bio": "STEM",
"history/holocaust & genocide studies": "Arts & Humanities",
"sped & eled": "Education", # Assuming this refers to special education & elementary education
"anthropology, communication studies": "Social Studies",
"theatre": "Arts & Humanities",
"studio art": "Arts & Humanities",
"urban planning and sustainable development": "Other", # Not a clear category, could be Social Studies or another category
"art and design": "Arts & Humanities",
"spanish": "Arts & Humanities", # Language studies are often classified here
"biochemistry": "STEM",
"art studio": "Arts & Humanities",
"art ed": "Arts & Humanities",
"comm": "Other", # Assuming "communication", but not explicitly listed, could fit Social Studies or Business,
"environmental studies: eco-social justice and education emphasis": "STEM", # Falls under Environmental Studies
"communications": "Other", # Often classified as Arts & Humanities
"theatre/education": "Education", # Falls under Education
"undecided": "Unknown", # Assuming still undecided as before
"marketing": "Business", # Falls under Business
"communication studies": "Arts & Humanities", # Often classified as Arts & Humanities
"sociology": "Social Studies", # Social Studies
"education and public relations": "Education", # Falls under Education
"pre nursing": "Health & Medicine", # Falls under Health & Medicine
"economics/mathematics": "STEM", # Combination of Economics and Mathematics - falls under STEM
"mathematics secondary education": "Education", # Falls under Education
"dance": "Arts & Humanities", # Often classified as Arts & Humanities
"art studio (ba), art history": "Arts & Humanities", # Falls under Arts & Humanities
"narrative and folklore studies (fairhaven major)": "Arts & Humanities", # Falls under Arts & Humanities
"pre med and psychology": "Health & Medicine", # Falls under Health & Medicine
"archaeology": "Social Studies", # Falls under Social Studies
"neuroscience": "STEM", # Falls under STEM
"english literature with a teaching emphasis": "Arts & Humanities", # Falls under Arts & Humanities
"marine science": "STEM", # Falls under STEM
"fairhaven": "Other", # Falls under Other
"international business": "Business", # Falls under Business
"music composition": "Arts & Humanities", # Falls under Arts & Humanities
"creative writing": "Arts & Humanities", # Falls under Arts & Humanities
"business and sustainability": "Business", # Falls under Business
"early childhood education": "Education", # Falls under Education
"graphic design": "Arts & Humanities", # Falls under Arts & Humanities
"education": "Education", # Falls under Education
"stem": "STEM", # Falls under STEM
"envs": "STEM", # Falls under STEM
"behavioural neuroscience": "STEM",
"english lit": "Arts & Humanities",
"food equity and sustainable agriculture": "Other", # Could be Social Studies, Business, or even STEM, unclear
"art history and museum studies": "Arts & Humanities",
"japanese language": "Arts & Humanities",
"graphic design and marketing": "Arts & Humanities", # Graphic Design is often in Arts & Humanities, Marketing is Business, but first seems primary
"music performance major": "Arts & Humanities",
"environment studies": "STEM",
"business or elementary education": "Unknown", # Could be either Business or Education
"marine and coastal science": "STEM",
"undeclared, strongly thinking about history": "Unknown", # Assuming lean towards Arts & Humanities, but no exact match
"public health": "Health & Medicine",
"energy policy and management": "Other", # Could be Business, Social Studies, or STEM
"undeclared": "Unknown",
"fine arts": "Arts & Humanities",
"english, history of culture": "Arts & Humanities",
"psychology and elementary education": "Education", # Both Psychology and Elementary Education could be Education
"communication science and disordwrs": "Health & Medicine",
"anthropology/history": "Social Studies",
"special education and elementary education": "Education",
"ibus": "Business", # Assuming International Business
"energy science": "STEM",
"politics/philosophy/economics": "Social Studies", # Combination of three Social Studies fields
"history/social studies": "Social Studies",
"energy": "STEM" # Energy could be a part of STEM disciplines like Physics or Environmental Sciences
}
noncategorized_data = []
def categorize_aos_string(aos):
if pd.isnull(aos) or aos == "n/a" or aos == "" or aos == "N/A":
return "Unknown" # For handling NaN values
field_clean = aos.lower().rstrip()
for key, category in manual_mapping_aos.items():
if key == field_clean:
return category
noncategorized_data.append(field_clean)
aos_df["AOSCat"] = aos_df["AOS"].apply(categorize_aos_string)
if (len(noncategorized_data) > 0):
raise ValueError(f"Unknown categories: {noncategorized_data}")
aos_df["AOSCat"].unique()
Out[16]:
array(['STEM', 'Other', 'Social Studies', 'Education',
'Arts & Humanities', 'Unknown', 'Business', 'Health & Medicine'],
dtype=object)
Shorting the Length of Single Select Answer Choices for "Style"¶
In [17]:
style_df = aos_df
style_df["Style"].unique()
Out[17]:
array(['Not interested in playing board/card games.',
'Party-only player, primarily playing board/card games at social gatherings or parties.',
'Strategy-focused player, enjoying games that require planning and tactics, but also participates in more casual games.',
'Situation-Specific Player, adapts style and enthusiasm based on the specific game or social context. May be casual in some situations and highly strategic in others, depending on the game being played.',
'None of these describe me.',
'Casual player, participating for fun and relaxation, without a strong focus on winning. Still willing to engage with complex games.',
'Simple and straightforward player, preferring uncomplicated games with easy rules.'],
dtype=object)
In [18]:
player_styles_mapping = {
"Strategy-focused player, enjoying games that require planning and tactics, but also participates in more casual games.": "Strategic",
"Situation-Specific Player, adapts style and enthusiasm based on the specific game or social context. May be casual in some situations and highly strategic in others, depending on the game being played.": "Situation-Specific",
"Casual player, participating for fun and relaxation, without a strong focus on winning. Still willing to engage with complex games.": "Casual",
"Simple and straightforward player, preferring uncomplicated games with easy rules.": "Simple",
"Party-only player, primarily playing board/card games at social gatherings or parties.": "Party/Social",
"Not interested in playing board/card games.": "Not Interested",
"Never played or never had the opportunity to play board/card games.": "Never Played",
"None of these describe me.": "Other",
"Prefer not to say": "Prefer not to say"
}
def map_player_styles(string):
for key, val in player_styles_mapping.items():
if key in string:
return val
raise ValueError(f"Unknown category: {string}")
new_style_df = style_df.copy()
new_style_df["Style"] = style_df["Style"].apply(map_player_styles)
new_style_df["Style"].unique()
Out[18]:
array(['Not Interested', 'Party/Social', 'Strategic',
'Situation-Specific', 'Other', 'Casual', 'Simple'], dtype=object)
Assigning Order to Particular Catagories¶
In [19]:
cat_order_df = new_style_df.copy()
cat_order_df["BoardGamesOwned"] = cat_order_df["BoardGamesOwned"].apply(lambda x: "0" if x == "I do not own any board/card games.\xa0" else x)
cat_order_df["BoardGamesOwned"] = pd.Categorical(cat_order_df["BoardGamesOwned"],
categories=["Prefer not to say","0", "1 or 2", "2 to 5", "5 to 10", "10 to 20", "More than 20"],ordered=True)
cat_order_df["FrequencyOfPlay"] = pd.Categorical(cat_order_df["FrequencyOfPlay"],
categories=["Daily","Several times a week","Weekly","Several times a month","Monthly","Every few months","Rarely/Seldom","Never","Prefer not to say",],ordered=True)
cat_order_df["ChessRating"] = cat_order_df["ChessRating"].apply(lambda x: "Unknown" if x == "" else x)
cat_order_df["ChessRating"] = pd.Categorical(cat_order_df["ChessRating"].astype(str),
categories=["Unknown","1","2","3","4","5","6","7","8","9","10"], ordered=True)
Convert Features to Int¶
In [20]:
convert_df = cat_order_df.copy()
convert_df["Age"] = convert_df["Age"].apply(lambda x: -1 if x == "" else x)
convert_df["Age"] = convert_df["Age"].astype(int)
In [21]:
cleaned_df = convert_df
cleaned_df.to_csv("datasets/cleaned.csv")
Exploratory Data Analysis¶
In [22]:
single_select_features = ['WWUStatus', 'Gender', 'Age', 'EmploymentStatus', 'Vision',
'Religiosity', 'AOSCat', 'EnjoysBoardGames', 'BoardGamesOwned',
'FrequencyOfPlay', 'Style', 'ChessRating']
race_bool_features = ['WhiteIsRace','AsianIsRace', 'BlackorAfricanAmericanIsRace', 'FilipinoIsRace',
'HispanicorLatinoIsRace','MixedethnicityIsRace', 'MultiracialIsRace', 'NativeAmericanorAmericanIndianIsRace',
'NativeHawaiianorPacificIslanderIsRace', 'PrefernottosayIsRace',]
element_bool_features = ['Conflict/CompetitionIsPreferredElement',
'CooperationIsPreferredElement',
'Heavy/ImmersiveThemingIsPreferredElement', 'LuckIsPreferredElement',
'Party/Low-StakesIsPreferredElement',
'Puzzle-SolvingIsPreferredElement',
'SocialDeduction/HiddenRoleIsPreferredElement',
'StrategyIsPreferredElement', 'TriviaIsPreferredElement',]
genre_bool_features = ['AbstractStrategyIsEnjoyedGenre', 'AdventureIsEnjoyedGenre',
'AnimalsIsEnjoyedGenre', 'AuctionIsEnjoyedGenre', 'CardIsEnjoyedGenre',
'CardDraftingIsEnjoyedGenre', 'CivilizationIsEnjoyedGenre',
'CooperativeIsEnjoyedGenre', 'Deck-BuildingIsEnjoyedGenre',
'DeductionIsEnjoyedGenre', 'EconomicIsEnjoyedGenre',
'EducationalIsEnjoyedGenre', 'ExplorationIsEnjoyedGenre',
'FantasyIsEnjoyedGenre', 'FarmingIsEnjoyedGenre',
'FightingIsEnjoyedGenre', 'HorrorIsEnjoyedGenre', 'LuckIsEnjoyedGenre',
'MedievalIsEnjoyedGenre', 'MemoryIsEnjoyedGenre',
'MiniaturesIsEnjoyedGenre', 'PartyIsEnjoyedGenre',
'PiratesIsEnjoyedGenre', 'PoliticalIsEnjoyedGenre',
'PuzzleIsEnjoyedGenre', 'RacingIsEnjoyedGenre',
'Role-PlayingIsEnjoyedGenre', 'RollandMoveIsEnjoyedGenre',
'ScienceFictionIsEnjoyedGenre',
'SocialDeduction/HiddenRoleIsEnjoyedGenre', 'SportsIsEnjoyedGenre',
'StrategyIsEnjoyedGenre', 'TerritoryBuildingIsEnjoyedGenre',
'Tile-LayingIsEnjoyedGenre', 'TrainsIsEnjoyedGenre',
'TransportationIsEnjoyedGenre', 'TravelIsEnjoyedGenre',
'TriviaIsEnjoyedGenre', 'WarIsEnjoyedGenre', 'WordIsEnjoyedGenre',
'WorkerPlacementIsEnjoyedGenre', 'WorldWarIIIsEnjoyedGenre',
'ZombiesIsEnjoyedGenre',]
free_form_features = ['AOS','EnjoyedBoardGames']
In [23]:
len(cleaned_df.columns)
Out[23]:
76
In [24]:
len(single_select_features + race_bool_features + element_bool_features + genre_bool_features + free_form_features)
Out[24]:
76
In [25]:
cleaned_df["Religiosity"].value_counts()
Out[25]:
Religiosity Atheism 59 No specific belief 55 Agnosticism 49 Christianity 32 Spiritual, not affiliated with a specific religion 31 Judaism 4 Prefer not to say 2 Pagan 1 Islam 1 Toaism 1 Lutheran 1 Hinduism 1 Buddhism 1 pagan 1 Unitarian 1 Science 1 being with oneself in connection to everything 1 Paganism 1 I believe a god exists but don’t follow any religious texts 1 Name: count, dtype: int64
In [26]:
religious_bucketing = {
"Christianity": "Religious",
"Judaism": "Religious",
"Islam": "Religious",
"Hinduism": "Religious",
"Buddhism": "Religious",
"Lutheran": "Religious",
"Pagan": "Religious",
"Paganism": "Religious",
"Unitarian": "Religious",
"I believe a god exists but don’t follow any religious texts": "Religious",
"Toaism": "Religious",
"pagan": "Religious",
"Atheism": "Not Religious",
"No specific belief": "Not Religious",
"Agnosticism": "Not Religious",
"Spiritual, not affiliated with a specific religion": "Not Religious",
"Science": "Not Religious",
"being with oneself in connection to everything": "Not Religious",
"Prefer not to say": "Not Religious",
}
Defining Functions for Analyzing Categorical Data¶
In [27]:
import matplotlib.pyplot as plt
import matplotlib.ticker as mtick
import seaborn as sns
import pandas as pd
def categories_against_category(df, categories, category):
for feature in categories:
print(df.groupby(category)[feature].value_counts())
# Compute percentage for each category within each cluster
df_percent = df.groupby(category)[feature].value_counts(normalize=True).rename('Percentage').reset_index()
df_percent["Percentage"] *= 100 # Convert to %
# Create seaborn barplot
plt.figure(figsize=(8, 6))
sns.barplot(x=feature, y='Percentage', hue=category, data=df_percent)
plt.gca().yaxis.set_major_formatter(mtick.PercentFormatter())
plt.xticks(rotation=90)
plt.show()
def bools_against_category(df,bools,category):
# Initialize an empty dataframe to store your percentages
percentages = pd.DataFrame()
# Loop through the bool_columns
for col in bools:
# Compute the percentages of True (=1) occurrences in each cluster
percents = df.groupby(category,observed=True)[col].mean().mul(100).reset_index()
percents['Feature'] = col
percents.rename({col: 'Percentage'}, axis=1, inplace=True)
# Append the computed percentages to your dataframe
percentages = pd.concat([percentages, percents])
# Once your dataframe is ready, you can plot it using seaborn
plt.figure(figsize=(12, 8))
sns.barplot(x='Feature', y='Percentage', hue=category, data=percentages)
plt.gca().yaxis.set_major_formatter(mtick.PercentFormatter())
plt.xticks(rotation=90)
plt.title(f"Percentage of 'True' Occurrences in Each Feature by {category}")
plt.show()
# Intended for single-select/categorical feature comparison
# `x` should be a Panda Series
# `y` should be a Panda Series
def plot_count_and_percentage_heatmaps(x, y):
data = pd.DataFrame({x.name: x, y.name: y})
count_matrix = pd.crosstab(data[x.name], data[y.name])
percent_matrix = count_matrix.div(count_matrix.sum(axis=1), axis=0) * 100
fig, ax = plt.subplots(1, 2, figsize=(12, 6))
# Heatmap for counts
sns.heatmap(count_matrix, annot=True, fmt="d", cmap="YlGnBu", cbar=True, ax=ax[0])
ax[0].set_title("Count Matrix of Categories")
ax[0].set_xlabel(y.name)
ax[0].set_ylabel(x.name)
# Heatmap for percentages
sns.heatmap(percent_matrix, annot=True, fmt=".2f", cmap="YlGnBu", cbar=True, ax=ax[1],vmin=0.0, vmax=100.0)
ax[1].set_title("Percentage Matrix of Categories")
ax[1].set_xlabel(y.name)
ax[1].set_ylabel(x.name)
plt.tight_layout()
plt.show()
# Intended for categorical feature comparison against several bool fields
# `data` should be a Pandas Data Frame
# `x` should be a string that refers to the categorical feature in `data`
# `y` should be an array of strings that refer to the several bool features in `data`
def plot_heatmap_of_bool_features_percent(data, x, y):
df_melted = data.melt(id_vars=x, value_vars=y, var_name="Bool", value_name="True")
df_pivot = df_melted.groupby([x,"Bool"]).sum().reset_index().pivot(index=x, columns="Bool", values="True")
total = data[x].value_counts()
df_pivot = df_pivot.join(total)
df_pivot.loc[:, df_pivot.columns != "count"] = ((df_pivot.loc[:, df_pivot.columns != "count"].div(df_pivot["count"], axis=0)) * 100).round(4)
print(df_pivot["count"])
df_pivot = df_pivot.drop(columns=["count"])
df_pivot = df_pivot.transpose()
sns.heatmap(df_pivot, annot=True, cmap="Blues", fmt=".2f",vmin=0.0, vmax=100.0)
# Intended for categorical feature comparison against several bool fields
# `data` should be a Pandas Data Frame
# `x` should be a string that refers to the categorical feature in `data`
# `y` should be an array of strings that refer to the several bool features in `data`
def plot_heatmap_of_bool_features_val_count(data, x, y):
df_melted = data.melt(id_vars=x, value_vars=y, var_name="Bool", value_name="True")
df_pivot = df_melted.groupby([x,"Bool"]).sum().reset_index().pivot(index=x, columns="Bool", values="True")
total = data[x].value_counts()
df_pivot = df_pivot.join(total)
df_pivot.rename(columns={"count": "UniqueMembersOfParticular" + x}, inplace=True)
df_pivot = df_pivot.transpose()
# Plot
sns.heatmap(df_pivot, annot=True, cmap="Blues", fmt=".2f")
# Used to reduce the number of categories in a feature by setting the value to "Other" for insufficient response quantity of specific catagories
# `df` should be a Pandas Data Frame
# `col_name` should be the string name of the column/feature you'd like to filter
# `threshold` should be the minimum count needed to maintain the categorical variable
# For example, if `Age` gets a bunch of responses between 18 and 24, and you get a few 40s and one 60. You can use this function to reduce the
# different categorical variables to simply 18 through 24 and "Other"
def filter_threshold(df, col_name, threshold):
df_copy = df.copy()
counts = df_copy[col_name].value_counts()
df_copy.loc[df_copy[col_name].isin(counts[counts < threshold].index), col_name] = "Other"
return df_copy
Distribution of Features¶
The following code is used to filter out feature responses that are underrepresented or minimal in survey data, thus preventing inaccurate conclusions that could arise from only looking at a small sample of a population. Most of features graphed below have at least two prominent catagories for which to perform data analysis, which is sufficient. This includes:
- Gender
- Age (within the range of 18 to 24)
- Employment Status (at least for Part-Time and Unemployed)
- Vision
- Area of Study Catagories
- Number of Board Games Owned
- Frequency of Play
- Style
- Chess Rating
However, a few fields did not get sufficient diverse responses to conclude anything, this includes:
- WWU Status
- Nearly everyone who took this survey was a student
- Religiosity
- Most people who filled out this survey were secular, and the largest minority religious group is simply too small to make any meaningful conclusions
- Does person enjoys board games
- The overwhelming majority of people who filled out this survey liked playing board games making relationships between demographics hard to draw
In [28]:
for feature in single_select_features:
sns.countplot(x=feature, data=cleaned_df)
plt.xticks(rotation=90)
plt.show()
Distribution of Features (cont'd)¶
The following questions were multi-select and resulted in bool fields. Because of this, we need to display the data slightly differently: totalling all "True"/selected values from the related features.
The multi-select questions were in regard to Race, Preferred Gameplay Elements, and Enjoyed Genres.
Preferred Gameplay Elements and Enjoyed Genres got a sufficient distribution of responses, however, Race did not receive sufficiently diverse responses for analysis.
In [29]:
true_counts = {}
for col in race_bool_features:
true_counts[col] = cleaned_df[cleaned_df[col] == 1].shape[0]
true_counts_series = pd.Series(true_counts)
sns.barplot(true_counts_series)
plt.xticks(rotation=90)
plt.show()
In [30]:
true_counts = {}
for col in element_bool_features:
true_counts[col] = cleaned_df[cleaned_df[col] == 1].shape[0]
true_counts_series = pd.Series(true_counts)
sns.barplot(true_counts_series)
plt.xticks(rotation=90)
plt.show()
In [31]:
true_counts = {}
for col in genre_bool_features:
true_counts[col] = cleaned_df[cleaned_df[col] == 1].shape[0]
true_counts_series = pd.Series(true_counts)
sns.barplot(true_counts_series)
plt.xticks(rotation=90)
plt.show()
(Random) Exploratory Analysis¶
This section is dedicated to graphing random sufficient features against one another in hopes of uncovering hidden relationships.
In [32]:
# There were many different responses for various types of non-binary gender identities, because of this, graphing the data becomes a bit muddled
# This is solved by grouping the non-binary responses into an "Other" category
genderForAnalysis_df = filter_threshold(cleaned_df,"Gender",25)
genderForAnalysis_df["Gender"].value_counts()
Out[32]:
Gender Woman 109 Man 101 Other 34 Name: count, dtype: int64
In [33]:
genderForAnalysis_df = genderForAnalysis_df[genderForAnalysis_df["Gender"] != "Other"]
In [34]:
categories_against_category(genderForAnalysis_df,single_select_features,"Gender")
Gender WWUStatus
Man WWU Student 98
WWU Faculty Member 2
Neither 1
Woman WWU Student 108
WWU Faculty Member 1
Name: count, dtype: int64
Gender Man 101 Woman 109 Name: count, dtype: int64
Gender Age
Man 18 28
19 23
20 16
21 11
22 6
23 6
-1 5
26 3
24 2
25 1
Woman 18 29
19 23
20 21
21 15
22 8
23 4
24 4
-1 1
25 1
28 1
29 1
36 1
Name: count, dtype: int64
Gender EmploymentStatus
Man Unemployed 51
Employed part-time 45
Employed full-time 2
Seasonal worker during the summer 1
Seeking employment 1
Self-employed 1
Woman Unemployed 53
Employed part-time 50
Prefer not to say 2
Employed full-time 1
One day a week babysitting 1
Seasonally employed 1
Self-employed 1
Name: count, dtype: int64
Gender Vision
Man None 61
Glasses 27
Both 7
Contacts 5
Prefer not to say 1
Woman None 55
Glasses 31
Both 18
Contacts 5
Name: count, dtype: int64
Gender Religiosity
Man Agnosticism 26
Atheism 25
No specific belief 24
Christianity 12
Spiritual, not affiliated with a specific religion 6
Judaism 3
Toaism 1
Science 1
Prefer not to say 1
Islam 1
Buddhism 1
Woman No specific belief 26
Atheism 22
Christianity 20
Agnosticism 19
Spiritual, not affiliated with a specific religion 17
I believe a god exists but don’t follow any religious texts 1
Lutheran 1
Judaism 1
Hinduism 1
Unitarian 1
Name: count, dtype: int64
Gender AOSCat
Man STEM 48
Arts & Humanities 16
Unknown 14
Social Studies 7
Business 6
Education 5
Other 3
Health & Medicine 2
Woman STEM 35
Arts & Humanities 21
Unknown 16
Education 10
Health & Medicine 8
Social Studies 8
Other 7
Business 4
Name: count, dtype: int64
Gender EnjoysBoardGames
Man Yes 93
No 8
Woman Yes 106
No 2
Prefer not to say 1
Name: count, dtype: int64
Gender BoardGamesOwned
Man 2 to 5 29
1 or 2 24
5 to 10 24
0 10
10 to 20 7
More than 20 7
Prefer not to say 0
Woman 5 to 10 34
2 to 5 31
1 or 2 27
10 to 20 9
0 4
More than 20 4
Prefer not to say 0
Name: count, dtype: int64
Gender FrequencyOfPlay
Man Several times a month 24
Every few months 18
Rarely/Seldom 17
Weekly 14
Monthly 13
Several times a week 10
Never 3
Daily 2
Prefer not to say 0
Woman Every few months 28
Several times a month 27
Monthly 21
Weekly 12
Rarely/Seldom 9
Several times a week 7
Daily 4
Never 1
Prefer not to say 0
Name: count, dtype: int64
Gender Style
Man Situation-Specific 38
Casual 31
Strategic 20
Party/Social 6
Not Interested 3
Simple 2
Other 1
Woman Situation-Specific 43
Casual 41
Strategic 15
Party/Social 6
Simple 4
Name: count, dtype: int64
Gender ChessRating
Man 7 25
8 17
6 12
4 11
3 8
5 8
1 8
9 4
10 4
2 4
Unknown 0
Woman 1 28
5 14
2 14
7 12
4 10
6 9
8 8
3 7
Unknown 3
10 3
9 1
Name: count, dtype: int64
In [35]:
bools_against_category(genderForAnalysis_df,race_bool_features,"Gender")
bools_against_category(genderForAnalysis_df,element_bool_features,"Gender")
bools_against_category(genderForAnalysis_df,genre_bool_features,"Gender")
All Single Select Against ChessRating Box Plot
There does appear to be a substantial difference in ratings between Men and Women within in the data.
In [36]:
genderForAnalysis_df["Gender"].value_counts()
Out[36]:
Gender Woman 109 Man 101 Name: count, dtype: int64
In [37]:
sns.histplot(x=cleaned_df["ChessRating"])
Out[37]:
<Axes: xlabel='ChessRating', ylabel='Count'>
In [38]:
modified_df = cleaned_df.copy()
modified_df = filter_threshold(modified_df,"Gender",30)
modified_df = filter_threshold(modified_df,"Age",10)
modified_df = filter_threshold(modified_df,"EmploymentStatus",30)
modified_df["Religiosity"] = modified_df["Religiosity"].map(religious_bucketing)
modified_df = filter_threshold(modified_df,"AOSCat",30)
modified_df["FrequencyOfPlay"] = modified_df["FrequencyOfPlay"].apply(lambda x: "At Least Weekly" if x in ["Daily","Several times a week","Weekly"] else x)
modified_df["FrequencyOfPlay"] = pd.Categorical(modified_df["FrequencyOfPlay"],
categories=["At Least Weekly","Several times a month","Monthly","Every few months","Rarely/Seldom","Never","Prefer not to say"],ordered=True)
features = single_select_features.copy()
features.remove("WWUStatus")
for feature in features:
print(modified_df[feature].value_counts())
plt.figure(figsize=(20, 10))
sns.boxenplot(x=modified_df[feature],y=modified_df["ChessRating"])
plt.ylim(reversed(plt.ylim()))
plt.xticks(rotation=90)
plt.show()
Gender Woman 109 Man 101 Other 34 Name: count, dtype: int64
Age 18 69 19 58 20 41 21 28 Other 23 22 15 23 10 Name: count, dtype: int64
EmploymentStatus Unemployed 124 Employed part-time 107 Other 13 Name: count, dtype: int64
Vision None 132 Glasses 76 Both 25 Contacts 10 Prefer not to say 1 Name: count, dtype: int64
Religiosity Not Religious 197 Religious 46 Name: count, dtype: int64
AOSCat STEM 95 Other 65 Arts & Humanities 47 Unknown 37 Name: count, dtype: int64
EnjoysBoardGames Yes 231 No 12 Prefer not to say 1 Name: count, dtype: int64
BoardGamesOwned 2 to 5 73 5 to 10 60 1 or 2 59 10 to 20 20 0 19 More than 20 13 Prefer not to say 0 Name: count, dtype: int64
FrequencyOfPlay Several times a month 62 At Least Weekly 56 Every few months 54 Monthly 37 Rarely/Seldom 30 Never 5 Prefer not to say 0 Name: count, dtype: int64
Style Situation-Specific 91 Casual 79 Strategic 42 Party/Social 19 Simple 8 Not Interested 4 Other 1 Name: count, dtype: int64
ChessRating 1 43 7 43 4 26 8 26 5 25 2 23 6 23 3 18 10 9 9 5 Unknown 3 Name: count, dtype: int64
In [39]:
# Concatenate the boolean columns with the "ChessRating" column
df_bool = cleaned_df[element_bool_features + ['ChessRating']]
# Reshape your DataFrame so that each boolean feature and its corresponding "ChessRating" are in a single row
df_melt = df_bool.melt(id_vars='ChessRating', var_name='Feature', value_name='Value')
# Select only the rows where the category is marked as True
df_melt = df_melt[df_melt['Value'] == 1]
# Now you can plot everything on the same plot
plt.figure(figsize=(20, 10))
sns.boxenplot(x='Feature', y='ChessRating', data=df_melt)
plt.ylim(reversed(plt.ylim()))
plt.xticks(rotation=90)
plt.title('Distribution of ChessRating for Each Feature')
plt.show()
In [40]:
# Concatenate the boolean columns with the "ChessRating" column
df_bool = cleaned_df[genre_bool_features + ['ChessRating']]
# Reshape your DataFrame so that each boolean feature and its corresponding "ChessRating" are in a single row
df_melt = df_bool.melt(id_vars='ChessRating', var_name='Feature', value_name='Value')
# Select only the rows where the category is marked as True
df_melt = df_melt[df_melt['Value'] == 1]
# Now you can plot everything on the same plot
plt.figure(figsize=(20, 10))
sns.boxenplot(x='Feature', y='ChessRating', data=df_melt)
plt.ylim(reversed(plt.ylim()))
plt.xticks(rotation=90)
plt.title('Distribution of ChessRating for Each Feature')
plt.show()
In [41]:
# This produces a bunch of plots, but it's hard to interpret as it is currently, and their is not an amazing amount of discovery
# modified_df = cleaned_df.copy()
# modified_df = filter_threshold(modified_df,"Gender",30)
# modified_df = filter_threshold(modified_df,"Age",10)
# modified_df = filter_threshold(modified_df,"EmploymentStatus",30)
# modified_df["Religiosity"] = modified_df["Religiosity"].map(religious_bucketing)
# modified_df = filter_threshold(modified_df,"AOSCat",30)
# features = single_select_features.copy()
# features.remove("WWUStatus")
# for feature in features:
# print(modified_df[feature].value_counts())
# bools_against_category(modified_df,element_bool_features,feature)
In [42]:
# This produces a bunch of plots, but it's hard to interpret as it is currently, and their is not an amazing amount of discovery
# modified_df = cleaned_df.copy()
# modified_df = filter_threshold(modified_df,"Gender",30)
# modified_df = filter_threshold(modified_df,"Age",10)
# modified_df = filter_threshold(modified_df,"EmploymentStatus",30)
# modified_df["Religiosity"] = modified_df["Religiosity"].map(religious_bucketing)
# modified_df = filter_threshold(modified_df,"AOSCat",30)
# features = single_select_features.copy()
# features.remove("WWUStatus")
# for feature in features:
# print(modified_df[feature].value_counts())
# bools_against_category(modified_df,genre_bool_features,feature)
In [43]:
cleaned_df["BoardGamesOwned"].value_counts()
Out[43]:
BoardGamesOwned 2 to 5 73 5 to 10 60 1 or 2 59 10 to 20 20 0 19 More than 20 13 Prefer not to say 0 Name: count, dtype: int64
In [44]:
selected_rows = cleaned_df.copy()
bools_against_category(selected_rows,element_bool_features,"BoardGamesOwned")
bools_against_category(selected_rows,genre_bool_features,"BoardGamesOwned")
In [45]:
modified_df = cleaned_df.copy()
modified_df = filter_threshold(modified_df,"Style",20)
modified_df = modified_df[modified_df["Style"] != "Other"]
print(modified_df["Style"].value_counts())
bools_against_category(modified_df,element_bool_features,"Style")
bools_against_category(modified_df,genre_bool_features,"Style")
Style Situation-Specific 91 Casual 79 Strategic 42 Name: count, dtype: int64
In [46]:
modified_df = cleaned_df.copy()
modified_df = filter_threshold(modified_df,"AOSCat",20)
print(modified_df["AOSCat"].value_counts())
features = single_select_features.copy()
features.remove("WWUStatus")
bools_against_category(modified_df,element_bool_features,"AOSCat")
bools_against_category(modified_df,genre_bool_features,"AOSCat")
AOSCat STEM 95 Other 65 Arts & Humanities 47 Unknown 37 Name: count, dtype: int64
Machine Learning Clustering for Exploratory Analysis¶
In [47]:
from sklearn.compose import ColumnTransformer
from sklearn.linear_model import LogisticRegression
from sklearn.neighbors import KNeighborsClassifier
from sklearn.ensemble import RandomForestClassifier
from sklearn.model_selection import train_test_split
from sklearn.pipeline import make_pipeline
from sklearn.preprocessing import OneHotEncoder, OrdinalEncoder
from sklearn.calibration import LabelEncoder
from sklearn.discriminant_analysis import StandardScaler
from sklearn.ensemble import RandomForestRegressor
from sklearn.metrics import confusion_matrix, accuracy_score, make_scorer, r2_score
from sklearn.model_selection import cross_val_score
from sklearn.tree import DecisionTreeClassifier
from sklearn.dummy import DummyClassifier
from sklearn import metrics
from sklearn.cluster import KMeans
data = cleaned_df.copy()
X = data.drop(columns=["EnjoyedBoardGames","AOS"])
onehot_and_ordinal_transform = ColumnTransformer(
transformers=[
("onehot", OneHotEncoder(), ["WWUStatus","EmploymentStatus","Vision","Religiosity","EnjoysBoardGames","Gender","Style","AOSCat"]),
("ordinal", make_pipeline(OrdinalEncoder(categories=[["Prefer not to say", "0", "1 or 2", "2 to 5", "5 to 10", "10 to 20", "More than 20"],
["Daily", "Several times a week", "Weekly", "Several times a month", "Monthly", "Every few months", "Rarely/Seldom", "Never", "Prefer not to say"],
["Unknown", "1", "2", "3", "4", "5", "6", "7", "8", "9", "10"]]),StandardScaler()),
["BoardGamesOwned",
"FrequencyOfPlay",
"ChessRating"]),
("scale numeric types", StandardScaler(), ["Age"])
], remainder="passthrough"
)
X_transformed = onehot_and_ordinal_transform.fit_transform(X)
for i in range(2,10):
k_means_model = KMeans(n_clusters=i, random_state=5, n_init=10)
clusters = k_means_model.fit_predict(X_transformed)
score = metrics.silhouette_score(X_transformed, clusters)
print(f'Fitting for {i} clusters')
print(f'score: {score}')
print()
Fitting for 2 clusters score: 0.11863466543320195 Fitting for 3 clusters score: 0.05927103971485946 Fitting for 4 clusters score: 0.053179399224910384 Fitting for 5 clusters score: 0.050326176098971546 Fitting for 6 clusters score: 0.053239186913892206 Fitting for 7 clusters score: 0.0531029232311401 Fitting for 8 clusters score: 0.049338481370941906 Fitting for 9 clusters score: 0.0474473146034939
In [48]:
k_means_model = KMeans(n_clusters=2, random_state=5, n_init=10)
clusters = k_means_model.fit_predict(X_transformed)
score = metrics.silhouette_score(X_transformed, clusters)
print(f'Fitting for {2} clusters')
print(f'score: {score}')
print()
Fitting for 2 clusters score: 0.11863466543320195
In [49]:
X["Cluster"] = k_means_model.fit_predict(X_transformed)
X
Out[49]:
| WWUStatus | Gender | Age | EmploymentStatus | Vision | Religiosity | EnjoysBoardGames | BoardGamesOwned | FrequencyOfPlay | Style | ... | TransportationIsEnjoyedGenre | TravelIsEnjoyedGenre | TriviaIsEnjoyedGenre | WarIsEnjoyedGenre | WordIsEnjoyedGenre | WorkerPlacementIsEnjoyedGenre | WorldWarIIIsEnjoyedGenre | ZombiesIsEnjoyedGenre | AOSCat | Cluster | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 0 | WWU Student | Man | 22 | Employed part-time | None | No specific belief | No | 0 | Never | Not Interested | ... | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | STEM | 0 |
| 1 | WWU Student | Man | 20 | Employed part-time | Contacts | Atheism | Yes | 2 to 5 | Rarely/Seldom | Party/Social | ... | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | STEM | 0 |
| 2 | WWU Student | Non-binary | 18 | Unemployed | None | Atheism | Yes | 10 to 20 | Several times a week | Strategic | ... | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | STEM | 0 |
| 3 | WWU Student | Man | 21 | Employed part-time | None | Atheism | Yes | 10 to 20 | Several times a month | Strategic | ... | 0 | 0 | 0 | 1 | 0 | 0 | 1 | 0 | STEM | 1 |
| 4 | WWU Student | Man | 22 | Employed part-time | None | Christianity | Yes | More than 20 | Weekly | Situation-Specific | ... | 0 | 0 | 1 | 1 | 1 | 1 | 1 | 1 | STEM | 1 |
| ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... |
| 239 | WWU Student | Man | 19 | Employed part-time | None | No specific belief | No | 2 to 5 | Rarely/Seldom | Situation-Specific | ... | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | STEM | 0 |
| 240 | WWU Student | Non-binary | 19 | Unemployed | Glasses | Agnosticism | Yes | 1 or 2 | Rarely/Seldom | Casual | ... | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | Social Studies | 0 |
| 241 | WWU Student | Non-binary | 19 | Employed part-time | None | Agnosticism | Yes | 2 to 5 | Several times a month | Situation-Specific | ... | 0 | 1 | 1 | 0 | 1 | 0 | 0 | 0 | STEM | 1 |
| 242 | WWU Student | Man | 18 | Unemployed | Glasses | Agnosticism | Yes | 5 to 10 | Several times a week | Casual | ... | 1 | 0 | 1 | 0 | 1 | 1 | 0 | 0 | Arts & Humanities | 1 |
| 243 | WWU Student | Man | 19 | Unemployed | Glasses | Agnosticism | Yes | 2 to 5 | Rarely/Seldom | Party/Social | ... | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | STEM | 0 |
244 rows × 75 columns
In [50]:
X["Cluster"].value_counts()
Out[50]:
Cluster 0 187 1 57 Name: count, dtype: int64
In [51]:
X["Cluster"].value_counts(normalize=True)
Out[51]:
Cluster 0 0.766393 1 0.233607 Name: proportion, dtype: float64
In [52]:
categories_against_category(X,single_select_features,"Cluster")
Cluster WWUStatus
0 WWU Student 184
WWU Faculty Member 3
1 WWU Student 56
Neither 1
Name: count, dtype: int64
Cluster Gender
0 Woman 82
Man 78
Non-binary 17
Genderfluid 3
Prefer not to say 2
Gender-fluid 1
Genderqueer 1
Wouldn't you like to know 1
girl thing 1
unsure 1
1 Woman 27
Man 23
Non-binary 6
Gender queer 1
Name: count, dtype: int64
Cluster Age
0 18 57
19 38
20 30
21 24
22 11
23 9
-1 7
24 4
26 3
25 2
36 1
28 1
1 19 20
18 12
20 11
21 4
22 4
24 2
26 1
23 1
-1 1
29 1
Name: count, dtype: int64
Cluster EmploymentStatus
0 Unemployed 92
Employed part-time 85
Employed full-time 3
Self-employed 3
One day a week babysitting 1
Prefer not to say 1
Seasonal worker during the summer 1
Seeking employment 1
1 Unemployed 32
Employed part-time 22
Employed full-time 1
Prefer not to say 1
Seasonally employed 1
Name: count, dtype: int64
Cluster Vision
0 None 108
Glasses 53
Both 16
Contacts 10
1 None 24
Glasses 23
Both 9
Prefer not to say 1
Name: count, dtype: int64
Cluster Religiosity
0 Atheism 46
No specific belief 40
Agnosticism 37
Spiritual, not affiliated with a specific religion 27
Christianity 22
Judaism 3
Prefer not to say 2
pagan 1
Science 1
being with oneself in connection to everything 1
Toaism 1
Paganism 1
Pagan 1
Lutheran 1
Islam 1
I believe a god exists but don’t follow any religious texts 1
Hinduism 1
1 No specific belief 15
Atheism 13
Agnosticism 12
Christianity 10
Spiritual, not affiliated with a specific religion 4
Judaism 1
Buddhism 1
Unitarian 1
Name: count, dtype: int64
Cluster AOSCat
0 STEM 74
Arts & Humanities 33
Unknown 26
Education 16
Social Studies 14
Business 8
Health & Medicine 8
Other 8
1 STEM 21
Arts & Humanities 14
Unknown 11
Social Studies 5
Business 2
Health & Medicine 2
Other 2
Name: count, dtype: int64
Cluster EnjoysBoardGames
0 Yes 175
No 11
Prefer not to say 1
1 Yes 56
No 1
Name: count, dtype: int64
Cluster BoardGamesOwned
0 2 to 5 62
1 or 2 54
5 to 10 40
0 19
10 to 20 8
More than 20 4
Prefer not to say 0
1 5 to 10 20
10 to 20 12
2 to 5 11
More than 20 9
1 or 2 5
Prefer not to say 0
0 0
Name: count, dtype: int64
Cluster FrequencyOfPlay
0 Every few months 44
Several times a month 43
Monthly 29
Rarely/Seldom 28
Weekly 23
Several times a week 10
Daily 5
Never 5
Prefer not to say 0
1 Several times a month 19
Every few months 10
Several times a week 9
Monthly 8
Weekly 7
Rarely/Seldom 2
Daily 2
Never 0
Prefer not to say 0
Name: count, dtype: int64
Cluster Style
0 Casual 68
Situation-Specific 54
Strategic 33
Party/Social 19
Simple 8
Not Interested 4
Other 1
1 Situation-Specific 37
Casual 11
Strategic 9
Name: count, dtype: int64
Cluster ChessRating
0 1 37
7 36
2 21
4 18
8 18
5 16
6 16
3 15
10 5
9 4
Unknown 1
1 5 9
4 8
8 8
6 7
7 7
1 6
10 4
3 3
Unknown 2
2 2
9 1
Name: count, dtype: int64
In [53]:
bools_against_category(X,race_bool_features,"Cluster")
bools_against_category(X,element_bool_features,"Cluster")
bools_against_category(X,genre_bool_features,"Cluster")
In [54]:
from sklearn.cluster import SpectralClustering
data = cleaned_df.copy()
X = data.drop(columns=["EnjoyedBoardGames","AOS","Age"])
onehot_and_ordinal_transform = ColumnTransformer(
transformers=[
("onehot", OneHotEncoder(), ["WWUStatus","EmploymentStatus","Vision","Religiosity","EnjoysBoardGames","Gender","Style","AOSCat"]),
("ordinal", make_pipeline(OrdinalEncoder(categories=[["Prefer not to say", "0", "1 or 2", "2 to 5", "5 to 10", "10 to 20", "More than 20"],
["Daily", "Several times a week", "Weekly", "Several times a month", "Monthly", "Every few months", "Rarely/Seldom", "Never", "Prefer not to say"],
["Unknown", "1", "2", "3", "4", "5", "6", "7", "8", "9", "10"]]),StandardScaler()),
["BoardGamesOwned",
"FrequencyOfPlay",
"ChessRating"])
], remainder="passthrough"
)
X_transformed = onehot_and_ordinal_transform.fit_transform(X)
for i in range(2,10):
spectral_clustering_model = SpectralClustering(n_clusters=i, random_state=5, n_init=10)
clusters = spectral_clustering_model.fit_predict(X_transformed)
score = metrics.silhouette_score(X_transformed, clusters)
print(f'Fitting for {i} clusters')
print(f'score: {score}')
print()
Fitting for 2 clusters score: 0.14680586026949505 Fitting for 3 clusters score: 0.10914232473605325 Fitting for 4 clusters score: 0.10934645867499984 Fitting for 5 clusters score: 0.02475783992583868 Fitting for 6 clusters score: 0.006370108937589488 Fitting for 7 clusters score: 0.00015076200052050723 Fitting for 8 clusters score: 0.048111530605883294 Fitting for 9 clusters score: 0.02940213489175527
In [55]:
spectral_clustering_model = SpectralClustering(n_clusters=2, random_state=5, n_init=10)
clusters = spectral_clustering_model.fit_predict(X_transformed)
score = metrics.silhouette_score(X_transformed, clusters)
print(f'Fitting for {2} clusters')
print(f'score: {score}')
print()
Fitting for 2 clusters score: 0.14680586026949505
In [56]:
X["Cluster"] = k_means_model.fit_predict(X_transformed)
X
Out[56]:
| WWUStatus | Gender | EmploymentStatus | Vision | Religiosity | EnjoysBoardGames | BoardGamesOwned | FrequencyOfPlay | Style | ChessRating | ... | TransportationIsEnjoyedGenre | TravelIsEnjoyedGenre | TriviaIsEnjoyedGenre | WarIsEnjoyedGenre | WordIsEnjoyedGenre | WorkerPlacementIsEnjoyedGenre | WorldWarIIIsEnjoyedGenre | ZombiesIsEnjoyedGenre | AOSCat | Cluster | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 0 | WWU Student | Man | Employed part-time | None | No specific belief | No | 0 | Never | Not Interested | 3 | ... | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | STEM | 0 |
| 1 | WWU Student | Man | Employed part-time | Contacts | Atheism | Yes | 2 to 5 | Rarely/Seldom | Party/Social | 6 | ... | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | STEM | 0 |
| 2 | WWU Student | Non-binary | Unemployed | None | Atheism | Yes | 10 to 20 | Several times a week | Strategic | 8 | ... | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | STEM | 0 |
| 3 | WWU Student | Man | Employed part-time | None | Atheism | Yes | 10 to 20 | Several times a month | Strategic | 9 | ... | 0 | 0 | 0 | 1 | 0 | 0 | 1 | 0 | STEM | 1 |
| 4 | WWU Student | Man | Employed part-time | None | Christianity | Yes | More than 20 | Weekly | Situation-Specific | 7 | ... | 0 | 0 | 1 | 1 | 1 | 1 | 1 | 1 | STEM | 1 |
| ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... |
| 239 | WWU Student | Man | Employed part-time | None | No specific belief | No | 2 to 5 | Rarely/Seldom | Situation-Specific | 6 | ... | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | STEM | 0 |
| 240 | WWU Student | Non-binary | Unemployed | Glasses | Agnosticism | Yes | 1 or 2 | Rarely/Seldom | Casual | 2 | ... | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | Social Studies | 0 |
| 241 | WWU Student | Non-binary | Employed part-time | None | Agnosticism | Yes | 2 to 5 | Several times a month | Situation-Specific | 5 | ... | 0 | 1 | 1 | 0 | 1 | 0 | 0 | 0 | STEM | 1 |
| 242 | WWU Student | Man | Unemployed | Glasses | Agnosticism | Yes | 5 to 10 | Several times a week | Casual | 8 | ... | 1 | 0 | 1 | 0 | 1 | 1 | 0 | 0 | Arts & Humanities | 1 |
| 243 | WWU Student | Man | Unemployed | Glasses | Agnosticism | Yes | 2 to 5 | Rarely/Seldom | Party/Social | 7 | ... | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | STEM | 0 |
244 rows × 74 columns
In [57]:
X["Cluster"].value_counts()
Out[57]:
Cluster 0 183 1 61 Name: count, dtype: int64
In [58]:
X["Cluster"].value_counts(normalize=True)
Out[58]:
Cluster 0 0.75 1 0.25 Name: proportion, dtype: float64
In [59]:
features = single_select_features.copy()
features.remove("Age")
categories_against_category(X,features,"Cluster")
Cluster WWUStatus
0 WWU Student 180
WWU Faculty Member 3
1 WWU Student 60
Neither 1
Name: count, dtype: int64
Cluster Gender
0 Woman 83
Man 74
Non-binary 16
Genderfluid 3
Prefer not to say 2
Gender-fluid 1
Genderqueer 1
Wouldn't you like to know 1
girl thing 1
unsure 1
1 Man 27
Woman 26
Non-binary 7
Gender queer 1
Name: count, dtype: int64
Cluster EmploymentStatus
0 Unemployed 89
Employed part-time 85
Self-employed 3
Employed full-time 2
One day a week babysitting 1
Prefer not to say 1
Seasonal worker during the summer 1
Seeking employment 1
1 Unemployed 35
Employed part-time 22
Employed full-time 2
Prefer not to say 1
Seasonally employed 1
Name: count, dtype: int64
Cluster Vision
0 None 107
Glasses 51
Both 15
Contacts 10
1 Glasses 25
None 25
Both 10
Prefer not to say 1
Name: count, dtype: int64
Cluster Religiosity
0 Atheism 44
No specific belief 40
Agnosticism 36
Spiritual, not affiliated with a specific religion 26
Christianity 22
Judaism 3
Prefer not to say 2
pagan 1
Science 1
being with oneself in connection to everything 1
Toaism 1
Paganism 1
Pagan 1
Lutheran 1
Islam 1
I believe a god exists but don’t follow any religious texts 1
Hinduism 1
1 No specific belief 15
Atheism 15
Agnosticism 13
Christianity 10
Spiritual, not affiliated with a specific religion 5
Judaism 1
Buddhism 1
Unitarian 1
Name: count, dtype: int64
Cluster AOSCat
0 STEM 71
Arts & Humanities 33
Unknown 26
Education 15
Social Studies 14
Business 8
Health & Medicine 8
Other 8
1 STEM 24
Arts & Humanities 14
Unknown 11
Social Studies 5
Business 2
Health & Medicine 2
Other 2
Education 1
Name: count, dtype: int64
Cluster EnjoysBoardGames
0 Yes 171
No 11
Prefer not to say 1
1 Yes 60
No 1
Name: count, dtype: int64
Cluster BoardGamesOwned
0 2 to 5 60
1 or 2 53
5 to 10 38
0 19
10 to 20 9
More than 20 4
Prefer not to say 0
1 5 to 10 22
2 to 5 13
10 to 20 11
More than 20 9
1 or 2 6
Prefer not to say 0
0 0
Name: count, dtype: int64
Cluster FrequencyOfPlay
0 Every few months 45
Several times a month 41
Monthly 28
Rarely/Seldom 27
Weekly 22
Several times a week 10
Daily 5
Never 5
Prefer not to say 0
1 Several times a month 21
Every few months 9
Several times a week 9
Monthly 9
Weekly 8
Rarely/Seldom 3
Daily 2
Never 0
Prefer not to say 0
Name: count, dtype: int64
Cluster Style
0 Casual 66
Situation-Specific 52
Strategic 33
Party/Social 19
Simple 8
Not Interested 4
Other 1
1 Situation-Specific 39
Casual 13
Strategic 9
Name: count, dtype: int64
Cluster ChessRating
0 1 36
7 34
2 21
4 17
5 17
8 17
6 16
3 15
10 5
9 4
Unknown 1
1 4 9
7 9
8 9
5 8
6 7
1 7
10 4
3 3
Unknown 2
2 2
9 1
Name: count, dtype: int64
In [60]:
bools_against_category(X,race_bool_features,"Cluster")
bools_against_category(X,element_bool_features,"Cluster")
bools_against_category(X,genre_bool_features,"Cluster")
Machine Learning Predictive Model¶
First Attempt at Machine Learning Model, Predicting Different Board Game Stats Based on Demographics¶
In [61]:
# load dataset
data = cleaned_df.copy()
data = data[['Gender', 'Age',
'AOSCat',
"BoardGamesOwned","FrequencyOfPlay","Style","ChessRating",]]
# data = data.drop(columns=["WWUStatus","EmploymentStatus","Vision","Religiosity","AOS","EnjoysBoardGames",
# "EnjoyedBoardGames"])
data = filter_threshold(data,"Gender",20)
# data = data[data["Gender"] != "Other"]
#filter_threshold(data,"Age",10)["Age"].value_counts()
# data = filter_threshold
data = filter_threshold(data,"AOSCat",15)
data = filter_threshold(data,"Style",15)
data["FrequencyOfPlay"] = data["FrequencyOfPlay"].apply(lambda x: "At Least Weekly" if x in ["Daily","Several times a week","Weekly"] else x)
data["FrequencyOfPlay"] = pd.Categorical(data["FrequencyOfPlay"],
categories=["At Least Weekly","Several times a month","Monthly","Every few months","Rarely/Seldom","Never","Prefer not to say"],ordered=True)
data["ChessRating"] = data["ChessRating"].replace('Unknown', np.nan).astype(float)
In [62]:
data = data.dropna(subset=['ChessRating'])
In [63]:
len(data)
Out[63]:
241
Age, Gender, and AOSCat cannot predict ChessRating
In [64]:
from sklearn.linear_model import LinearRegression
X, y = data.drop(columns=["BoardGamesOwned","FrequencyOfPlay","Style","ChessRating"]), data["ChessRating"]
X_train, X_test, y_train, y_test = train_test_split(X, y, train_size=0.5, test_size=0.5, stratify=y, random_state=10)
onehot_and_ordinal_transform = ColumnTransformer(
transformers=[
("onehot", OneHotEncoder(handle_unknown="ignore"), ["Gender","AOSCat", ]),
# ("ordinal", make_pipeline(OrdinalEncoder(categories=[["Prefer not to say", "0", "1 or 2", "2 to 5", "5 to 10", "10 to 20", "More than 20"],
# ["Daily", "Several times a week", "Weekly", "Several times a month", "Monthly", "Every few months", "Rarely/Seldom", "Never", "Prefer not to say"],
# ["Unknown", "1", "2", "3", "4", "5", "6", "7", "8", "9", "10"]]),StandardScaler()),
# ["BoardGamesOwned",
# "FrequencyOfPlay",
# "ChessRating"])
], remainder="passthrough"
)
linear_pipe = make_pipeline(
onehot_and_ordinal_transform,
LinearRegression()
)
dummy_classifier = DummyClassifier(strategy='most_frequent')
current_pipe = linear_pipe
print(f"Model Cross Val Scores: {cross_val_score(current_pipe,X_train,y_train,cv=3,scoring=make_scorer(r2_score))}")
print(f"Dummy Cross Val Scores: {cross_val_score(dummy_classifier,X_train,y_train,cv=3,scoring=make_scorer(r2_score))}")
Model Cross Val Scores: [-0.08918797 -0.0066436 -0.06956869] Dummy Cross Val Scores: [-1.94871795 -2.05991441 -1.99562437]
Age, Gender, and AOSCat cannot predict BoardGamesOwned
In [65]:
data["BoardGamesOwned"]
Out[65]:
0 0
1 2 to 5
2 10 to 20
3 10 to 20
4 More than 20
...
239 2 to 5
240 1 or 2
241 2 to 5
242 5 to 10
243 2 to 5
Name: BoardGamesOwned, Length: 241, dtype: category
Categories (7, object): ['Prefer not to say' < '0' < '1 or 2' < '2 to 5' < '5 to 10' < '10 to 20' < 'More than 20']
In [66]:
y_encoder = OrdinalEncoder()
y_translated = y_encoder.fit_transform(data["BoardGamesOwned"].values.reshape(-1, 1))
X, y = data.drop(columns=["BoardGamesOwned","FrequencyOfPlay","Style","ChessRating"]), y_translated
X_train, X_test, y_train, y_test = train_test_split(X, y, train_size=0.5, test_size=0.5, stratify=y, random_state=20)
onehot_and_ordinal_transform = ColumnTransformer(
transformers=[
("onehot", OneHotEncoder(handle_unknown="ignore"), ["Gender","AOSCat", ]),
# ("ordinal", make_pipeline(OrdinalEncoder(categories=[["Prefer not to say", "0", "1 or 2", "2 to 5", "5 to 10", "10 to 20", "More than 20"],
# ["Daily", "Several times a week", "Weekly", "Several times a month", "Monthly", "Every few months", "Rarely/Seldom", "Never", "Prefer not to say"],
# ["Unknown", "1", "2", "3", "4", "5", "6", "7", "8", "9", "10"]]),StandardScaler()),
# ["BoardGamesOwned",
# "FrequencyOfPlay",
# "ChessRating"])
], remainder="passthrough"
)
linear_pipe = make_pipeline(
onehot_and_ordinal_transform,
LinearRegression()
)
dummy_classifier = DummyClassifier(strategy='uniform')
current_pipe = linear_pipe
print(f"Model Cross Val Scores: {cross_val_score(current_pipe,X_train,y_train,cv=3,scoring=make_scorer(r2_score))}")
print(f"Dummy Cross Val Scores: {cross_val_score(dummy_classifier,X_train,y_train,cv=3,scoring=make_scorer(r2_score))}")
Model Cross Val Scores: [-0.27640396 -0.27075242 -0.2701293 ] Dummy Cross Val Scores: [-1.69770218 -1.72841051 -1.64364394]
In [67]:
current_pipe.fit(X_train,y_train)
y_encoder.inverse_transform(current_pipe.predict(X_train))
Out[67]:
array([['10 to 20'],
['10 to 20'],
['10 to 20'],
['10 to 20'],
['2 to 5'],
['10 to 20'],
['10 to 20'],
['10 to 20'],
['10 to 20'],
['10 to 20'],
['10 to 20'],
['10 to 20'],
['1 or 2'],
['2 to 5'],
['10 to 20'],
['10 to 20'],
['10 to 20'],
['10 to 20'],
['10 to 20'],
['10 to 20'],
['10 to 20'],
['10 to 20'],
['2 to 5'],
['10 to 20'],
['10 to 20'],
['10 to 20'],
['10 to 20'],
['2 to 5'],
['10 to 20'],
['2 to 5'],
['1 or 2'],
['1 or 2'],
['2 to 5'],
['10 to 20'],
['2 to 5'],
['10 to 20'],
['10 to 20'],
['10 to 20'],
['2 to 5'],
['10 to 20'],
['10 to 20'],
['10 to 20'],
['10 to 20'],
['1 or 2'],
['10 to 20'],
['10 to 20'],
['10 to 20'],
['10 to 20'],
['2 to 5'],
['10 to 20'],
['1 or 2'],
['10 to 20'],
['1 or 2'],
['10 to 20'],
['10 to 20'],
['10 to 20'],
['1 or 2'],
['10 to 20'],
['10 to 20'],
['10 to 20'],
['10 to 20'],
['1 or 2'],
['10 to 20'],
['10 to 20'],
['1 or 2'],
['10 to 20'],
['2 to 5'],
['1 or 2'],
['10 to 20'],
['1 or 2'],
['10 to 20'],
['10 to 20'],
['1 or 2'],
['10 to 20'],
['10 to 20'],
['10 to 20'],
['10 to 20'],
['1 or 2'],
['10 to 20'],
['2 to 5'],
['10 to 20'],
['10 to 20'],
['2 to 5'],
['1 or 2'],
['10 to 20'],
['10 to 20'],
['1 or 2'],
['10 to 20'],
['10 to 20'],
['10 to 20'],
['10 to 20'],
['10 to 20'],
['10 to 20'],
['10 to 20'],
['10 to 20'],
['1 or 2'],
['2 to 5'],
['10 to 20'],
['10 to 20'],
['10 to 20'],
['1 or 2'],
['10 to 20'],
['10 to 20'],
['10 to 20'],
['10 to 20'],
['10 to 20'],
['10 to 20'],
['1 or 2'],
['10 to 20'],
['10 to 20'],
['10 to 20'],
['10 to 20'],
['10 to 20'],
['10 to 20'],
['1 or 2'],
['10 to 20'],
['1 or 2'],
['10 to 20'],
['2 to 5'],
['10 to 20']], dtype=object)
In [68]:
y_encoder.inverse_transform(y_train)
Out[68]:
array([['2 to 5'],
['More than 20'],
['5 to 10'],
['5 to 10'],
['5 to 10'],
['1 or 2'],
['5 to 10'],
['2 to 5'],
['2 to 5'],
['10 to 20'],
['5 to 10'],
['5 to 10'],
['0'],
['5 to 10'],
['5 to 10'],
['5 to 10'],
['5 to 10'],
['2 to 5'],
['2 to 5'],
['1 or 2'],
['2 to 5'],
['1 or 2'],
['5 to 10'],
['2 to 5'],
['2 to 5'],
['2 to 5'],
['2 to 5'],
['More than 20'],
['0'],
['5 to 10'],
['1 or 2'],
['10 to 20'],
['5 to 10'],
['1 or 2'],
['10 to 20'],
['5 to 10'],
['2 to 5'],
['5 to 10'],
['2 to 5'],
['More than 20'],
['2 to 5'],
['1 or 2'],
['2 to 5'],
['0'],
['2 to 5'],
['10 to 20'],
['1 or 2'],
['5 to 10'],
['1 or 2'],
['10 to 20'],
['2 to 5'],
['2 to 5'],
['0'],
['5 to 10'],
['More than 20'],
['0'],
['2 to 5'],
['10 to 20'],
['0'],
['2 to 5'],
['5 to 10'],
['2 to 5'],
['1 or 2'],
['1 or 2'],
['1 or 2'],
['2 to 5'],
['2 to 5'],
['0'],
['1 or 2'],
['5 to 10'],
['2 to 5'],
['1 or 2'],
['0'],
['2 to 5'],
['1 or 2'],
['5 to 10'],
['2 to 5'],
['2 to 5'],
['5 to 10'],
['5 to 10'],
['1 or 2'],
['More than 20'],
['More than 20'],
['1 or 2'],
['1 or 2'],
['2 to 5'],
['1 or 2'],
['5 to 10'],
['1 or 2'],
['0'],
['5 to 10'],
['2 to 5'],
['1 or 2'],
['1 or 2'],
['10 to 20'],
['10 to 20'],
['2 to 5'],
['1 or 2'],
['1 or 2'],
['2 to 5'],
['5 to 10'],
['5 to 10'],
['More than 20'],
['1 or 2'],
['5 to 10'],
['2 to 5'],
['2 to 5'],
['1 or 2'],
['2 to 5'],
['1 or 2'],
['2 to 5'],
['5 to 10'],
['1 or 2'],
['1 or 2'],
['0'],
['5 to 10'],
['2 to 5'],
['1 or 2'],
['10 to 20'],
['2 to 5']], dtype=object)
In [69]:
current_pipe.fit(X_train,y_train)
current_pipe.score(X_train,y_train)
Out[69]:
0.11190381856394338
Age, Gender, and AOSCat cannot predict FrequencyOfPlay
In [70]:
y_encoder = OrdinalEncoder()
y_translated = y_encoder.fit_transform(data["FrequencyOfPlay"].values.reshape(-1, 1))
X, y = data.drop(columns=["BoardGamesOwned","FrequencyOfPlay","Style","ChessRating"]), y_translated
X_train, X_test, y_train, y_test = train_test_split(X, y, train_size=0.5, test_size=0.5, stratify=y, random_state=20)
onehot_and_ordinal_transform = ColumnTransformer(
transformers=[
("onehot", OneHotEncoder(handle_unknown="ignore"), ["Gender","AOSCat", ]),
# ("ordinal", make_pipeline(OrdinalEncoder(categories=[["Prefer not to say", "0", "1 or 2", "2 to 5", "5 to 10", "10 to 20", "More than 20"],
# ["Daily", "Several times a week", "Weekly", "Several times a month", "Monthly", "Every few months", "Rarely/Seldom", "Never", "Prefer not to say"],
# ["Unknown", "1", "2", "3", "4", "5", "6", "7", "8", "9", "10"]]),StandardScaler()),
# ["BoardGamesOwned",
# "FrequencyOfPlay",
# "ChessRating"])
], remainder="passthrough"
)
linear_pipe = make_pipeline(
onehot_and_ordinal_transform,
LinearRegression()
)
dummy_classifier = DummyClassifier(strategy='uniform')
current_pipe = linear_pipe
print(f"Model Cross Val Scores: {cross_val_score(current_pipe,X_train,y_train,cv=3,scoring=make_scorer(r2_score))}")
print(f"Dummy Cross Val Scores: {cross_val_score(dummy_classifier,X_train,y_train,cv=3,scoring=make_scorer(r2_score))}")
Model Cross Val Scores: [-0.22144966 -0.25985688 0.00776696] Dummy Cross Val Scores: [-0.78960872 -0.74718729 -0.55526142]
In [71]:
current_pipe.fit(X_train,y_train)
y_encoder.inverse_transform(current_pipe.predict(X_train))
Out[71]:
array([['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Every few months'],
['Every few months'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Every few months'],
['Monthly'],
['Monthly'],
['Monthly'],
['Never'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Every few months'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Every few months'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Every few months'],
['Monthly'],
['Every few months'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Every few months'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Every few months'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Rarely/Seldom'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Rarely/Seldom'],
['Never'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly'],
['Monthly']], dtype=object)
In [72]:
y_encoder.inverse_transform(y_train)
Out[72]:
array([['Monthly'],
['Several times a month'],
['Rarely/Seldom'],
['At Least Weekly'],
['Several times a month'],
['Several times a month'],
['Several times a month'],
['Monthly'],
['Every few months'],
['At Least Weekly'],
['At Least Weekly'],
['Several times a month'],
['Several times a month'],
['Every few months'],
['Monthly'],
['Several times a month'],
['Rarely/Seldom'],
['Rarely/Seldom'],
['At Least Weekly'],
['Several times a month'],
['Monthly'],
['Rarely/Seldom'],
['Every few months'],
['Monthly'],
['At Least Weekly'],
['Several times a month'],
['At Least Weekly'],
['Several times a month'],
['Every few months'],
['Rarely/Seldom'],
['Several times a month'],
['Monthly'],
['Rarely/Seldom'],
['Several times a month'],
['Rarely/Seldom'],
['Every few months'],
['Monthly'],
['At Least Weekly'],
['Every few months'],
['At Least Weekly'],
['Monthly'],
['Every few months'],
['Every few months'],
['Every few months'],
['Every few months'],
['Rarely/Seldom'],
['Every few months'],
['At Least Weekly'],
['Never'],
['Several times a month'],
['At Least Weekly'],
['Monthly'],
['Every few months'],
['At Least Weekly'],
['Monthly'],
['Several times a month'],
['Never'],
['At Least Weekly'],
['At Least Weekly'],
['Every few months'],
['Every few months'],
['Monthly'],
['At Least Weekly'],
['At Least Weekly'],
['Several times a month'],
['Rarely/Seldom'],
['At Least Weekly'],
['At Least Weekly'],
['At Least Weekly'],
['Rarely/Seldom'],
['Monthly'],
['Several times a month'],
['Rarely/Seldom'],
['At Least Weekly'],
['Several times a month'],
['Several times a month'],
['Every few months'],
['Every few months'],
['Monthly'],
['Every few months'],
['Several times a month'],
['Several times a month'],
['At Least Weekly'],
['Several times a month'],
['Several times a month'],
['Every few months'],
['At Least Weekly'],
['Every few months'],
['Every few months'],
['Rarely/Seldom'],
['At Least Weekly'],
['Every few months'],
['Never'],
['Several times a month'],
['Rarely/Seldom'],
['Several times a month'],
['Every few months'],
['Rarely/Seldom'],
['Several times a month'],
['Rarely/Seldom'],
['At Least Weekly'],
['Monthly'],
['At Least Weekly'],
['Several times a month'],
['Several times a month'],
['At Least Weekly'],
['At Least Weekly'],
['At Least Weekly'],
['Several times a month'],
['Monthly'],
['Every few months'],
['Every few months'],
['Monthly'],
['Several times a month'],
['Every few months'],
['Every few months'],
['Several times a month'],
['Monthly'],
['Monthly'],
['Several times a month']], dtype=object)
Second Attempt at Building a Machine Learning Model, Predict Gender from Board Game Preferences¶
In [73]:
# load dataset
data = cleaned_df.copy()
data = data.drop(columns=["WWUStatus","Age","EmploymentStatus","Vision","Religiosity","AOS","AOSCat","EnjoyedBoardGames"] + race_bool_features)
data = filter_threshold(data,"Gender",40)
data = data[data["Gender"] != "Other"]
# data = filter_threshold(data,"AOSCat",15)
data = filter_threshold(data,"Style",15)
In [74]:
X, y = data.drop(columns=["Gender"]), data["Gender"]
X_train, X_test, y_train, y_test = train_test_split(X, y, train_size=0.8, test_size=0.2, stratify=y, random_state=40)
onehot_and_ordinal_transform = ColumnTransformer(
transformers=[
("onehot", OneHotEncoder(handle_unknown="ignore"), ["Style","EnjoysBoardGames"]),
("ordinal", make_pipeline(OrdinalEncoder(categories=[["Prefer not to say", "0", "1 or 2", "2 to 5", "5 to 10", "10 to 20", "More than 20"],
["Daily", "Several times a week", "Weekly", "Several times a month", "Monthly", "Every few months", "Rarely/Seldom", "Never", "Prefer not to say"],
["Unknown", "1", "2", "3", "4", "5", "6", "7", "8", "9", "10"]]),StandardScaler()),
["BoardGamesOwned",
"FrequencyOfPlay",
"ChessRating"])
], remainder="passthrough"
)
logistic_pipe = make_pipeline(
onehot_and_ordinal_transform,
LogisticRegression(max_iter=10000)
)
k_neighbors_classifier_pipe = make_pipeline(
onehot_and_ordinal_transform,
KNeighborsClassifier()
)
decision_tree_classifier_pipe = make_pipeline(
onehot_and_ordinal_transform,
DecisionTreeClassifier()
)
random_forest_classifier_pipe = make_pipeline(
onehot_and_ordinal_transform,
RandomForestClassifier(random_state=40)
)
baseline_most_frequent = DummyClassifier(strategy='most_frequent')
baseline_uniform = DummyClassifier(strategy='uniform',random_state=40)
current_pipe = logistic_pipe
print(f"Model Cross Val Scores: {cross_val_score(current_pipe,X_train,y_train,cv=4,scoring=make_scorer(accuracy_score))}")
print(f"Dummy Cross Val Scores: {cross_val_score(baseline_most_frequent,X_train,y_train,cv=4,scoring=make_scorer(accuracy_score))}")
print(f"Dummy Cross Val Scores: {cross_val_score(baseline_uniform,X_train,y_train,cv=4,scoring=make_scorer(accuracy_score))}")
Model Cross Val Scores: [0.76190476 0.66666667 0.69047619 0.80952381] Dummy Cross Val Scores: [0.52380952 0.52380952 0.52380952 0.5 ] Dummy Cross Val Scores: [0.57142857 0.52380952 0.52380952 0.5 ]
In [75]:
current_pipe = k_neighbors_classifier_pipe
print(f"Model Cross Val Scores: {cross_val_score(current_pipe,X_train,y_train,cv=4,scoring=make_scorer(accuracy_score))}")
print(f"Dummy Cross Val Scores: {cross_val_score(baseline_most_frequent,X_train,y_train,cv=4,scoring=make_scorer(accuracy_score))}")
print(f"Dummy Cross Val Scores: {cross_val_score(baseline_uniform,X_train,y_train,cv=4,scoring=make_scorer(accuracy_score))}")
Model Cross Val Scores: [0.54761905 0.5952381 0.64285714 0.5952381 ] Dummy Cross Val Scores: [0.52380952 0.52380952 0.52380952 0.5 ] Dummy Cross Val Scores: [0.57142857 0.52380952 0.52380952 0.5 ]
In [76]:
current_pipe = decision_tree_classifier_pipe
print(f"Model Cross Val Scores: {cross_val_score(current_pipe,X_train,y_train,cv=4,scoring=make_scorer(accuracy_score))}")
print(f"Dummy Cross Val Scores: {cross_val_score(baseline_most_frequent,X_train,y_train,cv=4,scoring=make_scorer(accuracy_score))}")
print(f"Dummy Cross Val Scores: {cross_val_score(baseline_uniform,X_train,y_train,cv=4,scoring=make_scorer(accuracy_score))}")
Model Cross Val Scores: [0.61904762 0.47619048 0.5952381 0.5 ] Dummy Cross Val Scores: [0.52380952 0.52380952 0.52380952 0.5 ] Dummy Cross Val Scores: [0.57142857 0.52380952 0.52380952 0.5 ]
In [77]:
current_pipe = random_forest_classifier_pipe
print(f"Model Cross Val Scores: {cross_val_score(current_pipe,X_train,y_train,cv=4,scoring=make_scorer(accuracy_score))}")
print(f"Dummy Cross Val Scores: {cross_val_score(baseline_most_frequent,X_train,y_train,cv=4,scoring=make_scorer(accuracy_score))}")
print(f"Dummy Cross Val Scores: {cross_val_score(baseline_uniform,X_train,y_train,cv=4,scoring=make_scorer(accuracy_score))}")
Model Cross Val Scores: [0.5952381 0.57142857 0.69047619 0.71428571] Dummy Cross Val Scores: [0.52380952 0.52380952 0.52380952 0.5 ] Dummy Cross Val Scores: [0.57142857 0.52380952 0.52380952 0.5 ]
In [78]:
from sklearn.ensemble import RandomForestClassifier
from sklearn.model_selection import RandomizedSearchCV
# Number of trees in Random Forest
n_estimators = [int(x) for x in np.linspace(start = 200, stop = 2000, num = 10)]
# Maximum number of levels in tree
max_depth = [int(x) for x in np.linspace(10, 110, num = 11)]
max_depth.append(None)
# Minimum number of samples required to split a node
min_samples_split = [2, 5, 10]
# Minimum number of samples required at each leaf node
min_samples_leaf = [1, 2, 4]
# Method of selecting samples for training each tree
bootstrap = [True, False]
# Create the random grid
random_grid = {'n_estimators': n_estimators,
'max_depth': max_depth,
'min_samples_split': min_samples_split,
'min_samples_leaf': min_samples_leaf,
'bootstrap': bootstrap}
rf = RandomForestClassifier(random_state=40)
random_forest_randomized_search_cv = make_pipeline(
onehot_and_ordinal_transform,
RandomizedSearchCV(estimator = rf, param_distributions = random_grid, n_iter = 100, cv = 4, verbose=2, random_state=42, n_jobs = -1)
)
random_forest_randomized_search_cv.fit(X_train, y_train)
Fitting 4 folds for each of 100 candidates, totalling 400 fits [CV] END bootstrap=False, max_depth=None, min_samples_leaf=1, min_samples_split=2, n_estimators=2000; total time= 2.8s [CV] END bootstrap=True, max_depth=80, min_samples_leaf=2, min_samples_split=2, n_estimators=1800; total time= 3.1s [CV] END bootstrap=False, max_depth=None, min_samples_leaf=1, min_samples_split=2, n_estimators=800; total time= 1.1s [CV] END bootstrap=False, max_depth=10, min_samples_leaf=2, min_samples_split=2, n_estimators=800; total time= 1.1s [CV] END bootstrap=True, max_depth=90, min_samples_leaf=4, min_samples_split=2, n_estimators=1800; total time= 3.0s [CV] END bootstrap=False, max_depth=80, min_samples_leaf=4, min_samples_split=5, n_estimators=200; total time= 0.3s [CV] END bootstrap=False, max_depth=70, min_samples_leaf=2, min_samples_split=10, n_estimators=600; total time= 0.8s [CV] END bootstrap=False, max_depth=80, min_samples_leaf=1, min_samples_split=2, n_estimators=1600; total time= 2.2s [CV] END bootstrap=True, max_depth=110, min_samples_leaf=1, min_samples_split=10, n_estimators=1400; total time= 2.4s [CV] END bootstrap=False, max_depth=None, min_samples_leaf=4, min_samples_split=10, n_estimators=1600; total time= 2.1s [CV] END bootstrap=True, max_depth=80, min_samples_leaf=2, min_samples_split=5, n_estimators=1000; total time= 1.8s [CV] END bootstrap=False, max_depth=110, min_samples_leaf=2, min_samples_split=2, n_estimators=400; total time= 0.6s [CV] END bootstrap=True, max_depth=20, min_samples_leaf=1, min_samples_split=2, n_estimators=1400; total time= 2.5s [CV] END bootstrap=False, max_depth=10, min_samples_leaf=4, min_samples_split=10, n_estimators=1000; total time= 1.3s [CV] END bootstrap=True, max_depth=70, min_samples_leaf=1, min_samples_split=10, n_estimators=1600; total time= 2.8s [CV] END bootstrap=True, max_depth=20, min_samples_leaf=2, min_samples_split=5, n_estimators=1200; total time= 2.1s [CV] END bootstrap=True, max_depth=10, min_samples_leaf=1, min_samples_split=10, n_estimators=2000; total time= 3.4s [CV] END bootstrap=False, max_depth=70, min_samples_leaf=1, min_samples_split=10, n_estimators=400; total time= 0.5s [CV] END bootstrap=False, max_depth=90, min_samples_leaf=2, min_samples_split=5, n_estimators=1200; total time= 1.6s [CV] END bootstrap=False, max_depth=100, min_samples_leaf=2, min_samples_split=10, n_estimators=200; total time= 0.3s [CV] END bootstrap=True, max_depth=70, min_samples_leaf=4, min_samples_split=10, n_estimators=400; total time= 0.7s [CV] END bootstrap=False, max_depth=20, min_samples_leaf=2, min_samples_split=2, n_estimators=1000; total time= 1.3s [CV] END bootstrap=False, max_depth=50, min_samples_leaf=1, min_samples_split=5, n_estimators=1200; total time= 1.6s [CV] END bootstrap=True, max_depth=None, min_samples_leaf=4, min_samples_split=2, n_estimators=1600; total time= 2.7s [CV] END bootstrap=True, max_depth=40, min_samples_leaf=4, min_samples_split=10, n_estimators=200; total time= 0.3s [CV] END bootstrap=False, max_depth=60, min_samples_leaf=2, min_samples_split=10, n_estimators=1400; total time= 1.8s [CV] END bootstrap=True, max_depth=None, min_samples_leaf=1, min_samples_split=5, n_estimators=200; total time= 0.4s [CV] END bootstrap=True, max_depth=30, min_samples_leaf=4, min_samples_split=5, n_estimators=400; total time= 0.7s [CV] END bootstrap=False, max_depth=40, min_samples_leaf=4, min_samples_split=5, n_estimators=200; total time= 0.3s [CV] END bootstrap=False, max_depth=40, min_samples_leaf=4, min_samples_split=5, n_estimators=200; total time= 0.3s [CV] END bootstrap=True, max_depth=110, min_samples_leaf=4, min_samples_split=2, n_estimators=1200; total time= 2.0s [CV] END bootstrap=False, max_depth=20, min_samples_leaf=1, min_samples_split=5, n_estimators=2000; total time= 2.6s [CV] END bootstrap=False, max_depth=110, min_samples_leaf=1, min_samples_split=10, n_estimators=600; total time= 0.8s [CV] END bootstrap=True, max_depth=20, min_samples_leaf=1, min_samples_split=10, n_estimators=400; total time= 0.7s [CV] END bootstrap=True, max_depth=30, min_samples_leaf=1, min_samples_split=2, n_estimators=1000; total time= 1.7s [CV] END bootstrap=False, max_depth=80, min_samples_leaf=1, min_samples_split=5, n_estimators=1400; total time= 1.9s [CV] END bootstrap=False, max_depth=80, min_samples_leaf=2, min_samples_split=2, n_estimators=1600; total time= 2.1s [CV] END bootstrap=False, max_depth=70, min_samples_leaf=2, min_samples_split=5, n_estimators=1600; total time= 2.1s [CV] END bootstrap=False, max_depth=110, min_samples_leaf=2, min_samples_split=5, n_estimators=600; total time= 0.8s [CV] END bootstrap=True, max_depth=50, min_samples_leaf=4, min_samples_split=2, n_estimators=600; total time= 1.0s [CV] END bootstrap=True, max_depth=60, min_samples_leaf=2, min_samples_split=10, n_estimators=1800; total time= 3.2s [CV] END bootstrap=False, max_depth=30, min_samples_leaf=4, min_samples_split=5, n_estimators=400; total time= 0.5s [CV] END bootstrap=False, max_depth=50, min_samples_leaf=1, min_samples_split=10, n_estimators=2000; total time= 2.6s [CV] END bootstrap=False, max_depth=100, min_samples_leaf=2, min_samples_split=5, n_estimators=1200; total time= 1.6s [CV] END bootstrap=True, max_depth=80, min_samples_leaf=1, min_samples_split=2, n_estimators=200; total time= 0.4s [CV] END bootstrap=False, max_depth=40, min_samples_leaf=2, min_samples_split=5, n_estimators=800; total time= 1.0s [CV] END bootstrap=False, max_depth=100, min_samples_leaf=2, min_samples_split=5, n_estimators=800; total time= 1.0s [CV] END bootstrap=False, max_depth=30, min_samples_leaf=2, min_samples_split=2, n_estimators=800; total time= 1.1s [CV] END bootstrap=True, max_depth=110, min_samples_leaf=4, min_samples_split=5, n_estimators=800; total time= 1.3s [CV] END bootstrap=True, max_depth=40, min_samples_leaf=1, min_samples_split=10, n_estimators=1600; total time= 2.7s [CV] END bootstrap=False, max_depth=80, min_samples_leaf=2, min_samples_split=10, n_estimators=400; total time= 0.5s [CV] END bootstrap=True, max_depth=110, min_samples_leaf=1, min_samples_split=10, n_estimators=2000; total time= 3.4s [CV] END bootstrap=True, max_depth=60, min_samples_leaf=2, min_samples_split=5, n_estimators=1200; total time= 2.1s [CV] END bootstrap=False, max_depth=80, min_samples_leaf=2, min_samples_split=2, n_estimators=1000; total time= 1.3s [CV] END bootstrap=True, max_depth=40, min_samples_leaf=2, min_samples_split=2, n_estimators=1600; total time= 2.7s [CV] END bootstrap=True, max_depth=20, min_samples_leaf=2, min_samples_split=5, n_estimators=1000; total time= 1.7s [CV] END bootstrap=False, max_depth=60, min_samples_leaf=4, min_samples_split=10, n_estimators=1400; total time= 1.8s [CV] END bootstrap=False, max_depth=50, min_samples_leaf=4, min_samples_split=2, n_estimators=2000; total time= 2.6s [CV] END bootstrap=True, max_depth=50, min_samples_leaf=1, min_samples_split=2, n_estimators=400; total time= 0.7s [CV] END bootstrap=True, max_depth=20, min_samples_leaf=1, min_samples_split=5, n_estimators=2000; total time= 3.4s [CV] END bootstrap=False, max_depth=80, min_samples_leaf=2, min_samples_split=10, n_estimators=1200; total time= 1.5s [CV] END bootstrap=True, max_depth=110, min_samples_leaf=1, min_samples_split=10, n_estimators=800; total time= 1.3s [CV] END bootstrap=True, max_depth=30, min_samples_leaf=1, min_samples_split=2, n_estimators=1800; total time= 3.1s [CV] END bootstrap=False, max_depth=60, min_samples_leaf=2, min_samples_split=10, n_estimators=1000; total time= 1.3s [CV] END bootstrap=True, max_depth=30, min_samples_leaf=4, min_samples_split=2, n_estimators=200; total time= 0.3s [CV] END bootstrap=True, max_depth=30, min_samples_leaf=4, min_samples_split=2, n_estimators=200; total time= 0.4s [CV] END bootstrap=True, max_depth=20, min_samples_leaf=1, min_samples_split=5, n_estimators=200; total time= 0.3s [CV] END bootstrap=True, max_depth=70, min_samples_leaf=2, min_samples_split=10, n_estimators=1400; total time= 2.4s [CV] END bootstrap=True, max_depth=60, min_samples_leaf=2, min_samples_split=2, n_estimators=1400; total time= 2.5s [CV] END bootstrap=False, max_depth=None, min_samples_leaf=1, min_samples_split=2, n_estimators=2000; total time= 2.9s [CV] END bootstrap=True, max_depth=80, min_samples_leaf=2, min_samples_split=2, n_estimators=1800; total time= 3.1s [CV] END bootstrap=False, max_depth=None, min_samples_leaf=1, min_samples_split=2, n_estimators=800; total time= 1.1s [CV] END bootstrap=False, max_depth=10, min_samples_leaf=2, min_samples_split=2, n_estimators=800; total time= 1.1s [CV] END bootstrap=True, max_depth=90, min_samples_leaf=4, min_samples_split=2, n_estimators=1800; total time= 3.1s [CV] END bootstrap=False, max_depth=80, min_samples_leaf=4, min_samples_split=5, n_estimators=200; total time= 0.3s [CV] END bootstrap=False, max_depth=70, min_samples_leaf=2, min_samples_split=10, n_estimators=600; total time= 0.8s [CV] END bootstrap=False, max_depth=80, min_samples_leaf=1, min_samples_split=2, n_estimators=1600; total time= 2.2s [CV] END bootstrap=True, max_depth=110, min_samples_leaf=1, min_samples_split=10, n_estimators=1400; total time= 2.4s [CV] END bootstrap=False, max_depth=None, min_samples_leaf=4, min_samples_split=10, n_estimators=1600; total time= 2.2s [CV] END bootstrap=True, max_depth=80, min_samples_leaf=2, min_samples_split=5, n_estimators=1000; total time= 1.7s [CV] END bootstrap=False, max_depth=110, min_samples_leaf=2, min_samples_split=2, n_estimators=400; total time= 0.5s [CV] END bootstrap=True, max_depth=20, min_samples_leaf=1, min_samples_split=2, n_estimators=1400; total time= 2.4s [CV] END bootstrap=False, max_depth=10, min_samples_leaf=4, min_samples_split=10, n_estimators=1000; total time= 1.3s [CV] END bootstrap=True, max_depth=70, min_samples_leaf=1, min_samples_split=10, n_estimators=1600; total time= 2.7s [CV] END bootstrap=True, max_depth=20, min_samples_leaf=2, min_samples_split=5, n_estimators=1200; total time= 2.1s [CV] END bootstrap=True, max_depth=10, min_samples_leaf=1, min_samples_split=10, n_estimators=2000; total time= 3.4s [CV] END bootstrap=False, max_depth=90, min_samples_leaf=2, min_samples_split=5, n_estimators=1200; total time= 1.6s [CV] END bootstrap=False, max_depth=100, min_samples_leaf=2, min_samples_split=10, n_estimators=200; total time= 0.3s [CV] END bootstrap=False, max_depth=100, min_samples_leaf=2, min_samples_split=10, n_estimators=200; total time= 0.3s [CV] END bootstrap=True, max_depth=70, min_samples_leaf=4, min_samples_split=10, n_estimators=400; total time= 0.7s [CV] END bootstrap=False, max_depth=20, min_samples_leaf=2, min_samples_split=2, n_estimators=1000; total time= 1.3s [CV] END bootstrap=False, max_depth=50, min_samples_leaf=1, min_samples_split=5, n_estimators=1200; total time= 1.6s [CV] END bootstrap=True, max_depth=None, min_samples_leaf=4, min_samples_split=2, n_estimators=1600; total time= 2.7s [CV] END bootstrap=True, max_depth=40, min_samples_leaf=4, min_samples_split=10, n_estimators=200; total time= 0.4s [CV] END bootstrap=False, max_depth=60, min_samples_leaf=2, min_samples_split=10, n_estimators=1400; total time= 1.8s [CV] END bootstrap=True, max_depth=None, min_samples_leaf=1, min_samples_split=5, n_estimators=200; total time= 0.4s [CV] END bootstrap=True, max_depth=30, min_samples_leaf=4, min_samples_split=5, n_estimators=400; total time= 0.7s [CV] END bootstrap=False, max_depth=40, min_samples_leaf=4, min_samples_split=5, n_estimators=200; total time= 0.3s [CV] END bootstrap=True, max_depth=110, min_samples_leaf=4, min_samples_split=2, n_estimators=1200; total time= 2.0s [CV] END bootstrap=False, max_depth=20, min_samples_leaf=1, min_samples_split=5, n_estimators=2000; total time= 2.7s [CV] END bootstrap=False, max_depth=110, min_samples_leaf=1, min_samples_split=10, n_estimators=600; total time= 0.8s [CV] END bootstrap=True, max_depth=20, min_samples_leaf=1, min_samples_split=10, n_estimators=400; total time= 0.7s [CV] END bootstrap=True, max_depth=30, min_samples_leaf=1, min_samples_split=2, n_estimators=1000; total time= 1.7s [CV] END bootstrap=False, max_depth=80, min_samples_leaf=1, min_samples_split=5, n_estimators=1400; total time= 1.9s [CV] END bootstrap=False, max_depth=80, min_samples_leaf=2, min_samples_split=2, n_estimators=1600; total time= 2.1s [CV] END bootstrap=False, max_depth=70, min_samples_leaf=2, min_samples_split=5, n_estimators=1600; total time= 2.1s [CV] END bootstrap=False, max_depth=110, min_samples_leaf=2, min_samples_split=5, n_estimators=600; total time= 0.8s [CV] END bootstrap=True, max_depth=50, min_samples_leaf=4, min_samples_split=2, n_estimators=600; total time= 1.0s [CV] END bootstrap=True, max_depth=60, min_samples_leaf=2, min_samples_split=10, n_estimators=1800; total time= 3.1s [CV] END bootstrap=False, max_depth=30, min_samples_leaf=4, min_samples_split=5, n_estimators=400; total time= 0.5s [CV] END bootstrap=False, max_depth=50, min_samples_leaf=1, min_samples_split=10, n_estimators=2000; total time= 2.6s [CV] END bootstrap=False, max_depth=100, min_samples_leaf=2, min_samples_split=5, n_estimators=1200; total time= 1.6s [CV] END bootstrap=True, max_depth=80, min_samples_leaf=1, min_samples_split=2, n_estimators=200; total time= 0.4s [CV] END bootstrap=True, max_depth=80, min_samples_leaf=1, min_samples_split=2, n_estimators=200; total time= 0.4s [CV] END bootstrap=False, max_depth=40, min_samples_leaf=2, min_samples_split=5, n_estimators=800; total time= 1.1s [CV] END bootstrap=False, max_depth=100, min_samples_leaf=2, min_samples_split=5, n_estimators=800; total time= 1.1s [CV] END bootstrap=False, max_depth=30, min_samples_leaf=2, min_samples_split=2, n_estimators=800; total time= 1.1s [CV] END bootstrap=True, max_depth=110, min_samples_leaf=4, min_samples_split=5, n_estimators=800; total time= 1.4s [CV] END bootstrap=True, max_depth=40, min_samples_leaf=1, min_samples_split=10, n_estimators=1600; total time= 2.8s [CV] END bootstrap=False, max_depth=80, min_samples_leaf=2, min_samples_split=10, n_estimators=400; total time= 0.5s [CV] END bootstrap=True, max_depth=110, min_samples_leaf=1, min_samples_split=10, n_estimators=2000; total time= 3.4s [CV] END bootstrap=True, max_depth=60, min_samples_leaf=2, min_samples_split=5, n_estimators=1200; total time= 2.1s [CV] END bootstrap=False, max_depth=80, min_samples_leaf=2, min_samples_split=2, n_estimators=1000; total time= 1.3s [CV] END bootstrap=True, max_depth=40, min_samples_leaf=2, min_samples_split=2, n_estimators=1600; total time= 2.8s [CV] END bootstrap=True, max_depth=20, min_samples_leaf=2, min_samples_split=5, n_estimators=1000; total time= 1.7s [CV] END bootstrap=False, max_depth=60, min_samples_leaf=4, min_samples_split=10, n_estimators=1400; total time= 1.8s [CV] END bootstrap=False, max_depth=50, min_samples_leaf=4, min_samples_split=2, n_estimators=2000; total time= 2.6s [CV] END bootstrap=True, max_depth=50, min_samples_leaf=1, min_samples_split=2, n_estimators=400; total time= 0.7s [CV] END bootstrap=True, max_depth=20, min_samples_leaf=1, min_samples_split=5, n_estimators=2000; total time= 3.5s [CV] END bootstrap=False, max_depth=80, min_samples_leaf=2, min_samples_split=10, n_estimators=1200; total time= 1.6s [CV] END bootstrap=True, max_depth=110, min_samples_leaf=1, min_samples_split=10, n_estimators=800; total time= 1.4s [CV] END bootstrap=True, max_depth=30, min_samples_leaf=1, min_samples_split=2, n_estimators=1800; total time= 3.2s [CV] END bootstrap=False, max_depth=60, min_samples_leaf=2, min_samples_split=10, n_estimators=1000; total time= 1.3s [CV] END bootstrap=True, max_depth=20, min_samples_leaf=1, min_samples_split=5, n_estimators=200; total time= 0.3s [CV] END bootstrap=True, max_depth=70, min_samples_leaf=2, min_samples_split=10, n_estimators=1400; total time= 2.4s [CV] END bootstrap=True, max_depth=60, min_samples_leaf=2, min_samples_split=2, n_estimators=1400; total time= 2.4s [CV] END bootstrap=False, max_depth=50, min_samples_leaf=4, min_samples_split=2, n_estimators=1400; total time= 1.9s [CV] END bootstrap=False, max_depth=None, min_samples_leaf=1, min_samples_split=2, n_estimators=2000; total time= 2.8s [CV] END bootstrap=True, max_depth=80, min_samples_leaf=2, min_samples_split=2, n_estimators=1800; total time= 3.1s [CV] END bootstrap=False, max_depth=None, min_samples_leaf=1, min_samples_split=2, n_estimators=800; total time= 1.1s [CV] END bootstrap=False, max_depth=10, min_samples_leaf=2, min_samples_split=2, n_estimators=800; total time= 1.1s [CV] END bootstrap=True, max_depth=90, min_samples_leaf=4, min_samples_split=2, n_estimators=1800; total time= 3.1s [CV] END bootstrap=False, max_depth=80, min_samples_leaf=4, min_samples_split=5, n_estimators=200; total time= 0.3s [CV] END bootstrap=False, max_depth=70, min_samples_leaf=2, min_samples_split=10, n_estimators=600; total time= 0.8s [CV] END bootstrap=False, max_depth=80, min_samples_leaf=1, min_samples_split=2, n_estimators=1600; total time= 2.2s [CV] END bootstrap=True, max_depth=110, min_samples_leaf=1, min_samples_split=10, n_estimators=1400; total time= 2.4s [CV] END bootstrap=False, max_depth=None, min_samples_leaf=4, min_samples_split=10, n_estimators=1600; total time= 2.1s [CV] END bootstrap=True, max_depth=80, min_samples_leaf=2, min_samples_split=5, n_estimators=1000; total time= 1.7s [CV] END bootstrap=False, max_depth=110, min_samples_leaf=2, min_samples_split=2, n_estimators=400; total time= 0.5s [CV] END bootstrap=True, max_depth=20, min_samples_leaf=1, min_samples_split=2, n_estimators=1400; total time= 2.4s [CV] END bootstrap=False, max_depth=10, min_samples_leaf=4, min_samples_split=10, n_estimators=1000; total time= 1.3s [CV] END bootstrap=True, max_depth=70, min_samples_leaf=1, min_samples_split=10, n_estimators=1600; total time= 2.7s [CV] END bootstrap=True, max_depth=20, min_samples_leaf=2, min_samples_split=5, n_estimators=1200; total time= 2.0s [CV] END bootstrap=True, max_depth=10, min_samples_leaf=1, min_samples_split=10, n_estimators=2000; total time= 3.4s [CV] END bootstrap=False, max_depth=70, min_samples_leaf=1, min_samples_split=10, n_estimators=400; total time= 0.5s [CV] END bootstrap=False, max_depth=70, min_samples_leaf=1, min_samples_split=10, n_estimators=400; total time= 0.5s [CV] END bootstrap=False, max_depth=90, min_samples_leaf=2, min_samples_split=5, n_estimators=1200; total time= 1.6s [CV] END bootstrap=True, max_depth=70, min_samples_leaf=4, min_samples_split=10, n_estimators=400; total time= 0.7s [CV] END bootstrap=False, max_depth=20, min_samples_leaf=2, min_samples_split=2, n_estimators=1000; total time= 1.4s [CV] END bootstrap=False, max_depth=50, min_samples_leaf=1, min_samples_split=5, n_estimators=1200; total time= 1.6s [CV] END bootstrap=True, max_depth=None, min_samples_leaf=4, min_samples_split=2, n_estimators=1600; total time= 2.8s [CV] END bootstrap=True, max_depth=40, min_samples_leaf=4, min_samples_split=10, n_estimators=200; total time= 0.3s [CV] END bootstrap=False, max_depth=60, min_samples_leaf=2, min_samples_split=10, n_estimators=1400; total time= 1.9s [CV] END bootstrap=True, max_depth=None, min_samples_leaf=1, min_samples_split=5, n_estimators=200; total time= 0.3s [CV] END bootstrap=True, max_depth=30, min_samples_leaf=4, min_samples_split=5, n_estimators=400; total time= 0.7s [CV] END bootstrap=False, max_depth=40, min_samples_leaf=4, min_samples_split=5, n_estimators=200; total time= 0.3s [CV] END bootstrap=True, max_depth=110, min_samples_leaf=4, min_samples_split=2, n_estimators=1200; total time= 2.1s [CV] END bootstrap=False, max_depth=20, min_samples_leaf=1, min_samples_split=5, n_estimators=2000; total time= 2.7s [CV] END bootstrap=False, max_depth=110, min_samples_leaf=1, min_samples_split=10, n_estimators=600; total time= 0.8s [CV] END bootstrap=True, max_depth=20, min_samples_leaf=1, min_samples_split=10, n_estimators=400; total time= 0.7s [CV] END bootstrap=True, max_depth=30, min_samples_leaf=1, min_samples_split=2, n_estimators=1000; total time= 1.8s [CV] END bootstrap=False, max_depth=80, min_samples_leaf=1, min_samples_split=5, n_estimators=1400; total time= 1.9s [CV] END bootstrap=False, max_depth=80, min_samples_leaf=2, min_samples_split=2, n_estimators=1600; total time= 2.1s [CV] END bootstrap=False, max_depth=70, min_samples_leaf=2, min_samples_split=5, n_estimators=1600; total time= 2.1s [CV] END bootstrap=False, max_depth=110, min_samples_leaf=2, min_samples_split=5, n_estimators=600; total time= 0.8s [CV] END bootstrap=True, max_depth=50, min_samples_leaf=4, min_samples_split=2, n_estimators=600; total time= 1.0s [CV] END bootstrap=True, max_depth=60, min_samples_leaf=2, min_samples_split=10, n_estimators=1800; total time= 3.1s [CV] END bootstrap=False, max_depth=30, min_samples_leaf=4, min_samples_split=5, n_estimators=400; total time= 0.5s [CV] END bootstrap=False, max_depth=50, min_samples_leaf=1, min_samples_split=10, n_estimators=2000; total time= 2.7s [CV] END bootstrap=False, max_depth=100, min_samples_leaf=2, min_samples_split=5, n_estimators=1200; total time= 1.6s [CV] END bootstrap=True, max_depth=80, min_samples_leaf=1, min_samples_split=2, n_estimators=200; total time= 0.4s [CV] END bootstrap=False, max_depth=40, min_samples_leaf=2, min_samples_split=5, n_estimators=800; total time= 1.1s [CV] END bootstrap=False, max_depth=100, min_samples_leaf=2, min_samples_split=5, n_estimators=800; total time= 1.1s [CV] END bootstrap=False, max_depth=30, min_samples_leaf=2, min_samples_split=2, n_estimators=800; total time= 1.1s [CV] END bootstrap=True, max_depth=110, min_samples_leaf=4, min_samples_split=5, n_estimators=800; total time= 1.4s [CV] END bootstrap=True, max_depth=40, min_samples_leaf=1, min_samples_split=10, n_estimators=1600; total time= 2.8s [CV] END bootstrap=False, max_depth=80, min_samples_leaf=2, min_samples_split=10, n_estimators=400; total time= 0.5s [CV] END bootstrap=True, max_depth=110, min_samples_leaf=1, min_samples_split=10, n_estimators=2000; total time= 3.4s [CV] END bootstrap=True, max_depth=60, min_samples_leaf=2, min_samples_split=5, n_estimators=1200; total time= 2.1s [CV] END bootstrap=False, max_depth=80, min_samples_leaf=2, min_samples_split=2, n_estimators=1000; total time= 1.4s [CV] END bootstrap=True, max_depth=40, min_samples_leaf=2, min_samples_split=2, n_estimators=1600; total time= 2.8s [CV] END bootstrap=True, max_depth=20, min_samples_leaf=2, min_samples_split=5, n_estimators=1000; total time= 1.7s [CV] END bootstrap=False, max_depth=60, min_samples_leaf=4, min_samples_split=10, n_estimators=1400; total time= 1.8s [CV] END bootstrap=False, max_depth=50, min_samples_leaf=4, min_samples_split=2, n_estimators=2000; total time= 2.6s [CV] END bootstrap=True, max_depth=50, min_samples_leaf=1, min_samples_split=2, n_estimators=400; total time= 0.7s [CV] END bootstrap=True, max_depth=20, min_samples_leaf=1, min_samples_split=5, n_estimators=2000; total time= 3.5s [CV] END bootstrap=False, max_depth=80, min_samples_leaf=2, min_samples_split=10, n_estimators=1200; total time= 1.6s [CV] END bootstrap=True, max_depth=110, min_samples_leaf=1, min_samples_split=10, n_estimators=800; total time= 1.4s [CV] END bootstrap=True, max_depth=30, min_samples_leaf=1, min_samples_split=2, n_estimators=1800; total time= 3.1s [CV] END bootstrap=False, max_depth=60, min_samples_leaf=2, min_samples_split=10, n_estimators=1000; total time= 1.4s [CV] END bootstrap=True, max_depth=20, min_samples_leaf=1, min_samples_split=5, n_estimators=200; total time= 0.3s [CV] END bootstrap=True, max_depth=70, min_samples_leaf=2, min_samples_split=10, n_estimators=1400; total time= 2.4s [CV] END bootstrap=True, max_depth=60, min_samples_leaf=2, min_samples_split=2, n_estimators=1400; total time= 2.4s [CV] END bootstrap=False, max_depth=50, min_samples_leaf=4, min_samples_split=2, n_estimators=1400; total time= 1.8s [CV] END bootstrap=False, max_depth=90, min_samples_leaf=4, min_samples_split=2, n_estimators=1800; total time= 2.3s [CV] END bootstrap=False, max_depth=None, min_samples_leaf=1, min_samples_split=2, n_estimators=2000; total time= 2.9s [CV] END bootstrap=True, max_depth=80, min_samples_leaf=2, min_samples_split=2, n_estimators=1800; total time= 3.1s [CV] END bootstrap=False, max_depth=None, min_samples_leaf=1, min_samples_split=2, n_estimators=800; total time= 1.1s [CV] END bootstrap=False, max_depth=10, min_samples_leaf=2, min_samples_split=2, n_estimators=800; total time= 1.1s [CV] END bootstrap=True, max_depth=90, min_samples_leaf=4, min_samples_split=2, n_estimators=1800; total time= 3.0s [CV] END bootstrap=False, max_depth=80, min_samples_leaf=4, min_samples_split=5, n_estimators=200; total time= 0.3s [CV] END bootstrap=False, max_depth=70, min_samples_leaf=2, min_samples_split=10, n_estimators=600; total time= 0.8s [CV] END bootstrap=False, max_depth=80, min_samples_leaf=1, min_samples_split=2, n_estimators=1600; total time= 2.2s [CV] END bootstrap=True, max_depth=110, min_samples_leaf=1, min_samples_split=10, n_estimators=1400; total time= 2.5s [CV] END bootstrap=False, max_depth=None, min_samples_leaf=4, min_samples_split=10, n_estimators=1600; total time= 2.1s [CV] END bootstrap=True, max_depth=80, min_samples_leaf=2, min_samples_split=5, n_estimators=1000; total time= 1.7s [CV] END bootstrap=False, max_depth=110, min_samples_leaf=2, min_samples_split=2, n_estimators=400; total time= 0.5s [CV] END bootstrap=True, max_depth=20, min_samples_leaf=1, min_samples_split=2, n_estimators=1400; total time= 2.4s [CV] END bootstrap=False, max_depth=10, min_samples_leaf=4, min_samples_split=10, n_estimators=1000; total time= 1.4s [CV] END bootstrap=True, max_depth=70, min_samples_leaf=1, min_samples_split=10, n_estimators=1600; total time= 2.7s [CV] END bootstrap=True, max_depth=20, min_samples_leaf=2, min_samples_split=5, n_estimators=1200; total time= 2.1s [CV] END bootstrap=True, max_depth=10, min_samples_leaf=1, min_samples_split=10, n_estimators=2000; total time= 3.5s [CV] END bootstrap=False, max_depth=70, min_samples_leaf=1, min_samples_split=10, n_estimators=400; total time= 0.5s [CV] END bootstrap=False, max_depth=90, min_samples_leaf=2, min_samples_split=5, n_estimators=1200; total time= 1.6s [CV] END bootstrap=False, max_depth=100, min_samples_leaf=2, min_samples_split=10, n_estimators=200; total time= 0.3s [CV] END bootstrap=True, max_depth=70, min_samples_leaf=4, min_samples_split=10, n_estimators=400; total time= 0.7s [CV] END bootstrap=False, max_depth=20, min_samples_leaf=2, min_samples_split=2, n_estimators=1000; total time= 1.3s [CV] END bootstrap=False, max_depth=50, min_samples_leaf=1, min_samples_split=5, n_estimators=1200; total time= 1.6s [CV] END bootstrap=True, max_depth=None, min_samples_leaf=4, min_samples_split=2, n_estimators=1600; total time= 2.7s [CV] END bootstrap=True, max_depth=40, min_samples_leaf=4, min_samples_split=10, n_estimators=200; total time= 0.3s [CV] END bootstrap=False, max_depth=60, min_samples_leaf=2, min_samples_split=10, n_estimators=1400; total time= 1.8s [CV] END bootstrap=True, max_depth=None, min_samples_leaf=1, min_samples_split=5, n_estimators=200; total time= 0.4s [CV] END bootstrap=True, max_depth=30, min_samples_leaf=4, min_samples_split=5, n_estimators=400; total time= 0.7s [CV] END bootstrap=True, max_depth=110, min_samples_leaf=4, min_samples_split=2, n_estimators=1200; total time= 2.2s [CV] END bootstrap=False, max_depth=20, min_samples_leaf=1, min_samples_split=5, n_estimators=2000; total time= 2.9s [CV] END bootstrap=False, max_depth=110, min_samples_leaf=1, min_samples_split=10, n_estimators=600; total time= 0.8s [CV] END bootstrap=True, max_depth=20, min_samples_leaf=1, min_samples_split=10, n_estimators=400; total time= 0.7s [CV] END bootstrap=True, max_depth=30, min_samples_leaf=1, min_samples_split=2, n_estimators=1000; total time= 1.8s [CV] END bootstrap=False, max_depth=80, min_samples_leaf=1, min_samples_split=5, n_estimators=1400; total time= 1.9s [CV] END bootstrap=False, max_depth=80, min_samples_leaf=2, min_samples_split=2, n_estimators=1600; total time= 2.2s [CV] END bootstrap=False, max_depth=70, min_samples_leaf=2, min_samples_split=5, n_estimators=1600; total time= 2.2s [CV] END bootstrap=False, max_depth=110, min_samples_leaf=2, min_samples_split=5, n_estimators=600; total time= 0.8s [CV] END bootstrap=True, max_depth=50, min_samples_leaf=4, min_samples_split=2, n_estimators=600; total time= 1.0s [CV] END bootstrap=True, max_depth=60, min_samples_leaf=2, min_samples_split=10, n_estimators=1800; total time= 3.1s [CV] END bootstrap=False, max_depth=30, min_samples_leaf=4, min_samples_split=5, n_estimators=400; total time= 0.5s [CV] END bootstrap=False, max_depth=50, min_samples_leaf=1, min_samples_split=10, n_estimators=2000; total time= 2.6s [CV] END bootstrap=False, max_depth=100, min_samples_leaf=2, min_samples_split=5, n_estimators=1200; total time= 1.6s [CV] END bootstrap=False, max_depth=40, min_samples_leaf=2, min_samples_split=5, n_estimators=800; total time= 1.0s [CV] END bootstrap=False, max_depth=100, min_samples_leaf=2, min_samples_split=5, n_estimators=800; total time= 1.1s [CV] END bootstrap=False, max_depth=30, min_samples_leaf=2, min_samples_split=2, n_estimators=800; total time= 1.1s [CV] END bootstrap=True, max_depth=110, min_samples_leaf=4, min_samples_split=5, n_estimators=800; total time= 1.4s [CV] END bootstrap=True, max_depth=40, min_samples_leaf=1, min_samples_split=10, n_estimators=1600; total time= 2.7s [CV] END bootstrap=False, max_depth=80, min_samples_leaf=2, min_samples_split=10, n_estimators=400; total time= 0.6s [CV] END bootstrap=True, max_depth=110, min_samples_leaf=1, min_samples_split=10, n_estimators=2000; total time= 3.4s [CV] END bootstrap=True, max_depth=60, min_samples_leaf=2, min_samples_split=5, n_estimators=1200; total time= 2.0s [CV] END bootstrap=False, max_depth=80, min_samples_leaf=2, min_samples_split=2, n_estimators=1000; total time= 1.3s [CV] END bootstrap=True, max_depth=40, min_samples_leaf=2, min_samples_split=2, n_estimators=1600; total time= 2.8s [CV] END bootstrap=True, max_depth=20, min_samples_leaf=2, min_samples_split=5, n_estimators=1000; total time= 1.7s [CV] END bootstrap=False, max_depth=60, min_samples_leaf=4, min_samples_split=10, n_estimators=1400; total time= 1.8s [CV] END bootstrap=False, max_depth=50, min_samples_leaf=4, min_samples_split=2, n_estimators=2000; total time= 2.7s [CV] END bootstrap=True, max_depth=50, min_samples_leaf=1, min_samples_split=2, n_estimators=400; total time= 0.7s [CV] END bootstrap=True, max_depth=20, min_samples_leaf=1, min_samples_split=5, n_estimators=2000; total time= 3.4s [CV] END bootstrap=False, max_depth=80, min_samples_leaf=2, min_samples_split=10, n_estimators=1200; total time= 1.6s [CV] END bootstrap=True, max_depth=110, min_samples_leaf=1, min_samples_split=10, n_estimators=800; total time= 1.4s [CV] END bootstrap=True, max_depth=30, min_samples_leaf=1, min_samples_split=2, n_estimators=1800; total time= 3.1s [CV] END bootstrap=False, max_depth=60, min_samples_leaf=2, min_samples_split=10, n_estimators=1000; total time= 1.3s [CV] END bootstrap=True, max_depth=30, min_samples_leaf=4, min_samples_split=2, n_estimators=200; total time= 0.3s [CV] END bootstrap=True, max_depth=30, min_samples_leaf=4, min_samples_split=2, n_estimators=200; total time= 0.3s [CV] END bootstrap=True, max_depth=20, min_samples_leaf=1, min_samples_split=5, n_estimators=200; total time= 0.3s [CV] END bootstrap=True, max_depth=70, min_samples_leaf=2, min_samples_split=10, n_estimators=1400; total time= 2.5s [CV] END bootstrap=True, max_depth=60, min_samples_leaf=2, min_samples_split=2, n_estimators=1400; total time= 2.5s [CV] END bootstrap=False, max_depth=50, min_samples_leaf=4, min_samples_split=2, n_estimators=1400; total time= 1.8s [CV] END bootstrap=False, max_depth=90, min_samples_leaf=4, min_samples_split=2, n_estimators=1800; total time= 2.3s
Out[78]:
Pipeline(steps=[('columntransformer',
ColumnTransformer(remainder='passthrough',
transformers=[('onehot',
OneHotEncoder(handle_unknown='ignore'),
['Style',
'EnjoysBoardGames']),
('ordinal',
Pipeline(steps=[('ordinalencoder',
OrdinalEncoder(categories=[['Prefer '
'not '
'to '
'say',
'0',
'1 '
'or '
'2',
'2 '
'to '
'5',
'5 '
'to '
'10',
'10 '
'to '
'20',
'More '
'than '
'20'],
['Daily',
'Sev...
('randomizedsearchcv',
RandomizedSearchCV(cv=4,
estimator=RandomForestClassifier(random_state=40),
n_iter=100, n_jobs=-1,
param_distributions={'bootstrap': [True,
False],
'max_depth': [10, 20,
30, 40,
50, 60,
70, 80,
90, 100,
110,
None],
'min_samples_leaf': [1,
2,
4],
'min_samples_split': [2,
5,
10],
'n_estimators': [200,
400,
600,
800,
1000,
1200,
1400,
1600,
1800,
2000]},
random_state=42, verbose=2))])In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.
Pipeline(steps=[('columntransformer',
ColumnTransformer(remainder='passthrough',
transformers=[('onehot',
OneHotEncoder(handle_unknown='ignore'),
['Style',
'EnjoysBoardGames']),
('ordinal',
Pipeline(steps=[('ordinalencoder',
OrdinalEncoder(categories=[['Prefer '
'not '
'to '
'say',
'0',
'1 '
'or '
'2',
'2 '
'to '
'5',
'5 '
'to '
'10',
'10 '
'to '
'20',
'More '
'than '
'20'],
['Daily',
'Sev...
('randomizedsearchcv',
RandomizedSearchCV(cv=4,
estimator=RandomForestClassifier(random_state=40),
n_iter=100, n_jobs=-1,
param_distributions={'bootstrap': [True,
False],
'max_depth': [10, 20,
30, 40,
50, 60,
70, 80,
90, 100,
110,
None],
'min_samples_leaf': [1,
2,
4],
'min_samples_split': [2,
5,
10],
'n_estimators': [200,
400,
600,
800,
1000,
1200,
1400,
1600,
1800,
2000]},
random_state=42, verbose=2))])ColumnTransformer(remainder='passthrough',
transformers=[('onehot',
OneHotEncoder(handle_unknown='ignore'),
['Style', 'EnjoysBoardGames']),
('ordinal',
Pipeline(steps=[('ordinalencoder',
OrdinalEncoder(categories=[['Prefer '
'not '
'to '
'say',
'0',
'1 '
'or '
'2',
'2 '
'to '
'5',
'5 '
'to '
'10',
'10 '
'to '
'20',
'More '
'than '
'20'],
['Daily',
'Several '
'times '
'a '
'week',
'Weekly',
'Several '
'times '
'a '
'month',
'Monthly',
'Every '
'few '
'months',
'Rarely/Seldom',
'Never',
'Prefer '
'not '
'to '
'say'],
['Unknown',
'1',
'2',
'3',
'4',
'5',
'6',
'7',
'8',
'9',
'10']])),
('standardscaler',
StandardScaler())]),
['BoardGamesOwned', 'FrequencyOfPlay',
'ChessRating'])])['Style', 'EnjoysBoardGames']
OneHotEncoder(handle_unknown='ignore')
['BoardGamesOwned', 'FrequencyOfPlay', 'ChessRating']
OrdinalEncoder(categories=[['Prefer not to say', '0', '1 or 2', '2 to 5',
'5 to 10', '10 to 20', 'More than 20'],
['Daily', 'Several times a week', 'Weekly',
'Several times a month', 'Monthly',
'Every few months', 'Rarely/Seldom', 'Never',
'Prefer not to say'],
['Unknown', '1', '2', '3', '4', '5', '6', '7', '8',
'9', '10']])StandardScaler()
['Conflict/CompetitionIsPreferredElement', 'CooperationIsPreferredElement', 'Heavy/ImmersiveThemingIsPreferredElement', 'LuckIsPreferredElement', 'Party/Low-StakesIsPreferredElement', 'Puzzle-SolvingIsPreferredElement', 'SocialDeduction/HiddenRoleIsPreferredElement', 'StrategyIsPreferredElement', 'TriviaIsPreferredElement', 'AbstractStrategyIsEnjoyedGenre', 'AdventureIsEnjoyedGenre', 'AnimalsIsEnjoyedGenre', 'AuctionIsEnjoyedGenre', 'CardIsEnjoyedGenre', 'CardDraftingIsEnjoyedGenre', 'CivilizationIsEnjoyedGenre', 'CooperativeIsEnjoyedGenre', 'Deck-BuildingIsEnjoyedGenre', 'DeductionIsEnjoyedGenre', 'EconomicIsEnjoyedGenre', 'EducationalIsEnjoyedGenre', 'ExplorationIsEnjoyedGenre', 'FantasyIsEnjoyedGenre', 'FarmingIsEnjoyedGenre', 'FightingIsEnjoyedGenre', 'HorrorIsEnjoyedGenre', 'LuckIsEnjoyedGenre', 'MedievalIsEnjoyedGenre', 'MemoryIsEnjoyedGenre', 'MiniaturesIsEnjoyedGenre', 'PartyIsEnjoyedGenre', 'PiratesIsEnjoyedGenre', 'PoliticalIsEnjoyedGenre', 'PuzzleIsEnjoyedGenre', 'RacingIsEnjoyedGenre', 'Role-PlayingIsEnjoyedGenre', 'RollandMoveIsEnjoyedGenre', 'ScienceFictionIsEnjoyedGenre', 'SocialDeduction/HiddenRoleIsEnjoyedGenre', 'SportsIsEnjoyedGenre', 'StrategyIsEnjoyedGenre', 'TerritoryBuildingIsEnjoyedGenre', 'Tile-LayingIsEnjoyedGenre', 'TrainsIsEnjoyedGenre', 'TransportationIsEnjoyedGenre', 'TravelIsEnjoyedGenre', 'TriviaIsEnjoyedGenre', 'WarIsEnjoyedGenre', 'WordIsEnjoyedGenre', 'WorkerPlacementIsEnjoyedGenre', 'WorldWarIIIsEnjoyedGenre', 'ZombiesIsEnjoyedGenre']
passthrough
RandomizedSearchCV(cv=4, estimator=RandomForestClassifier(random_state=40),
n_iter=100, n_jobs=-1,
param_distributions={'bootstrap': [True, False],
'max_depth': [10, 20, 30, 40, 50, 60,
70, 80, 90, 100, 110,
None],
'min_samples_leaf': [1, 2, 4],
'min_samples_split': [2, 5, 10],
'n_estimators': [200, 400, 600, 800,
1000, 1200, 1400, 1600,
1800, 2000]},
random_state=42, verbose=2)RandomForestClassifier(random_state=40)
RandomForestClassifier(random_state=40)
In [79]:
cv_results = random_forest_randomized_search_cv.named_steps['randomizedsearchcv'].cv_results_
In [80]:
# for mean_score, params in zip(cv_results["mean_test_score"], cv_results["params"]):
# print(params, 'has a cross-validated score of', mean_score)
In [81]:
random_forest_randomized_search_cv.named_steps['randomizedsearchcv'].best_params_
Out[81]:
{'n_estimators': 2000,
'min_samples_split': 10,
'min_samples_leaf': 1,
'max_depth': 110,
'bootstrap': True}
In [82]:
random_forest_randomized_search_cv.named_steps['randomizedsearchcv'].best_score_
Out[82]:
0.7202380952380953
In [83]:
conf_matrix = confusion_matrix(y_test, random_forest_randomized_search_cv.predict(X_test),labels=["Man","Woman",])
print(conf_matrix)
print(f"Model Accuracy Score: {accuracy_score(random_forest_randomized_search_cv.predict(X_test),y_test)}")
baseline_most_frequent.fit(X_train, y_train)
y_pred_most_frequent = baseline_most_frequent.predict(X_test)
baseline_uniform.fit(X_train,y_train)
y_pred_uniform = baseline_uniform.predict(X_test)
print(f"Baseline Most Frequent Score: {accuracy_score(y_test, y_pred_most_frequent)}")
print(f"Baseline Uniform Score: {accuracy_score(y_test, y_pred_uniform)}")
[[15 5] [ 4 18]] Model Accuracy Score: 0.7857142857142857 Baseline Most Frequent Score: 0.5238095238095238 Baseline Uniform Score: 0.42857142857142855
In [84]:
logistic_pipe.fit(X_train,y_train)
conf_matrix = confusion_matrix(y_test, logistic_pipe.predict(X_test),labels=["Man","Woman"])
print(conf_matrix)
print(f"Model Accuracy Score: {accuracy_score(logistic_pipe.predict(X_test),y_test)}")
baseline_most_frequent.fit(X_train, y_train)
y_pred_most_frequent = baseline_most_frequent.predict(X_test)
baseline_uniform.fit(X_train,y_train)
y_pred_uniform = baseline_uniform.predict(X_test)
print(f"Baseline Most Frequent Score: {accuracy_score(y_test, y_pred_most_frequent)}")
print(f"Baseline Uniform Score: {accuracy_score(y_test, y_pred_uniform)}")
[[12 8] [ 5 17]] Model Accuracy Score: 0.6904761904761905 Baseline Most Frequent Score: 0.5238095238095238 Baseline Uniform Score: 0.42857142857142855